# paper.pdf ## Metadata - PDFFormatVersion=1.4 - Language=en-US - IsLinearized=false - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - Author= Maïtena Saurel - CreationDate=D:20260513161101+08'00' - Creator=Adobe InDesign 15.1 (Windows) - Custom.CrossMarkDomains[1]=springer.com - Custom.CrossMarkDomains[2]=springerlink.com - Custom.CrossmarkDomainExclusive=true - Custom.CrossmarkMajorVersionDate=2010-04-23 - Custom.doi=10.1007/s44470-026-00079-7 - Custom.robots=noindex - Keywords=Intraindividual variability;Sleep irregularity;Sleep variability;Accelerometry;Sleepiness;Long covid - ModDate=D:20260513171749+02'00' - Producer=Adobe PDF Library 15.0; modified using iText® 5.5.13.4 ©2000-2024 iText Group NV (SPRINGER SBM; licensed version) - Subject=Journal of Clinical Sleep Medicine, https://doi.org/10.1007/s44470-026-00079-7 - Title=Sleep in myalgic encephalomyelitis/chronic fatigue syndrome shows marked night-to-night fluctuation under free-living conditions—results from a matched case-control study - Trapped=False - dc:publisher=Springer International Publishing - dc:description=Journal of Clinical Sleep Medicine, https://doi.org/10.1007/s44470-026-00079-7 - dc:subject=Intraindividual variability,Sleep irregularity,Sleep variability,Accelerometry,Sleepiness,Long covid - dc:title=Sleep in myalgic encephalomyelitis/chronic fatigue syndrome shows marked night-to-night fluctuation under free-living conditions—results from a matched case-control study - dc:creator=Maïtena Saurel,Isabelle Fornasieri,Giovanna C. Del Sordo,Cyril Chatain,Maria Livia Fantini,Mathieu Gruet,Oussama Saidi - crossmark:crossmarkdomains=springer.comspringerlink.com - pdfx:crossmarkdomains=springer.comspringerlink.com ## Contents ### Page 1 Vol.:(0123456789)Journal of Clinical Sleep Medicine (2026) 22:77 https://doi.org/10.1007/s44470-026-00079-7ORIGINAL ARTICLESleep in myalgic encephalomyelitis/chronic fatigue syndrome shows marked night‑to‑night fluctuation under free‑living conditions—results from a matched case‑control studyMaïtena Saurel 1 · Isabelle Fornasieri2,3 · Giovanna C. Del Sordo 4 · Cyril Chatain 1,5 · Maria Livia Fantini6 · Mathieu Gruet1 · Oussama Saidi1Received: 18 January 2026 / Revised: 12 March 2026 / Accepted: 24 March 2026 © The Author(s) 2026AbstractPurpose Unrefreshing and non-restorative sleep is a hallmark complaint in people with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). However, little is known about their habitual sleep and night-to-night fluctuations under real- life conditions. This study aimed to characterize sleep, and the intraindividual variability (IIV) of sleep in people living with ME/CFS compared with matched controls.Methods In this case-control study, 38 ME/CFS and 38 controls wore a wrist accelerometer continuously for 7 days and completed concurrent sleep diaries, the Pittsburgh Sleep Quality Index (PSQI), and Epworth Sleepiness Scale (ESS). Within the ME/CFS group, participants were also stratified by symptom severity using the Bell Disability Scale. Sleep IIV was quantified using the coefficient of variation, the root mean square of successive differences, and the Bayesian variability model, respectively.Results Compared with controls, individuals with ME/CFS spent significantly more time in bed and exhibited poorer sleep efficiency (SE) (all p < 0.05). Despite a longer time in bed, total sleep time did not differ between groups. ME/CFS par- ticipants also displayed significantly greater IIV in SE. By contrast, sleep timing (bedtime) was more regular among ME/ CFS. Exploratory analyses did not detect clear differences across ME/CFS severity subgroups for mean sleep variables or variability indices.Conclusion Under real-life conditions, people with ME/CFS exhibit poor sleep quality and unstable SE. These findings highlight sleep IIV as a clinically relevant dimension of sleep health in ME/CFS.Brief summaryCurrent knowledge/study rationale Unrefreshing sleep is a core symptom of ME/CFS, yet most evidence relies on single- or two-night laboratory assessments that may not reflect habitual sleep under real-life conditions. Moreover, night-to-night sleep variability, a potentially critical dimension of sleep health, has not been systematically examined in ME/CFS.Study impact Using week-long wrist accelerometry, this study shows that under free-living conditions sleep in ME/CFS is characterized not only by impaired sleep efficiency but also by pronounced night-to-night variability, despite relatively stable bedtime compared to controls. These findings highlight sleep efficiency variability as a clinically relevant feature of ME/CFS and underscore the need for multi-night assessment and targeted strategies addressing sleep variability.* Oussama Saidioussama.saidi@univ-tln.fr 1 Laboratory Youth-Physical Activity, Sport and Health (J-AP2S), University of Toulon, CS 60584, 83041, Cedex 9 Toulon, France 2 ASFC, French ME/CFS Association, Nice, France 3 Faculty of Psychology, University of Strasbourg, Strasbourg, France 4 Psychology Department, New Mexico State University, 1780 E University Blvd, Las Cruces, NM 88003, USA 5 Inserm, CAPS UMR 1093, Burgundy Europe University, 21000 Dijon, France 6 Sleep Disorders Center, Neurophysiology Unit, Neurology Department, Clermont-Ferrand University Hospital, Université Clermont Auvergne, CNRS, Institut Pascal, 58, Rue Montalembert, 63000 Clermont-Ferrand, France ### Page 2 Journal of Clinical Sleep Medicine (2026) 22:7777 Page 2 of 13Keywords Intraindividual variability · Sleep irregularity · Sleep variability · Accelerometry · Sleepiness · Long covidAbbreviationsCFS Chronic fatigue syndrome COVID Coronavirus disease CV Coefficient of variation ESS Epworth Sleepiness Scale GGIR Generic GitHub Implementation of Raw Accelerometer Analysis in R ICC International Consensus Criteria IIV Intraindividual variability IOM Institute of Medicine IRB Institutional Review Board ME/CFS Myalgic encephalomyelitis/chronic fatigue syndrome PSG Polysomnography PSQI Pittsburgh Sleep Quality Index REM Rapid eye movement SARS-CoV-2 Severe acute respiratory syndrome corona- virus 2 SE Sleep efficiency SOL Sleep onset latency TIB Time in bed TST Total sleep time WASO Wake after sleep onsetIntroductionMyalgic encephalomyelitis/chronic fatigue syndrome (ME/ CFS) is a complex, multifactorial condition characterized by persistent fatigue and a variety of symptoms that drastically decrease daily activity levels and quality of life [1, 2]. Although the physiopathology of this chronic condition remains largely unknown, it appears to involve widespread abnormalities affecting the central nervous system, the immune system, and the endocrine system, as well as metabolism [3–5]. The true extent of ME/CFS global impact remains debated; however, it is estimated to affect between 10 and 70 million individu- als worldwide, exceeding the prevalence of conditions such as multiple sclerosis or lupus [6, 7]. ME/CFS is frequently triggered by several viral infections, such as Epstein-Barr virus and other herpes viruses [8]. Recently, SARS-CoV-2 has been identified as a potential trigger, with nearly half of individuals with long COVID meeting the clinical diagnostic criteria for ME/CFS [9, 10]. This overlap has given rise to a resurgence of interest in ME/CFS, pointing out the necessity for a more profound comprehension of its symptoms, and the develop- ment of more effective management strategies [11]. As outlined in the 2015 Institute of Medicine (IOM) diagnostic criteria, unrefreshing sleep is a key symptom of ME/CFS [12]. A recent systematic review and meta-analysis reported increased time in bed (TIB) and total sleep time (TST) in people with ME/CFS compared with their healthy counterparts [13]. Despite the prolonged sleep duration, peo- ple with ME/CFS exhibited altered sleep quality, as indi- cated by decreased sleep efficiency (SE), increased sleep onset latency (SOL), and wake after sleep onset (WASO). Abnormalities in sleep staging were also reported, with a significant reduction in N2 and an increase in N3 sleep proportion. Although the reduction in rapid eye movement (REM) sleep did not reach significance, the meta-analysis documented a valuable increase in REM sleep latency. The majority of studies in this area were based on polysomno- graphic sleep assessment (PSG) (16 out of 20 studies in adults). Although PSG remains the gold standard strategy for diagnosing primary sleep disorders and examining sleep architecture, one or two nights of sleep assessment may not be representative of habitual sleep and may not provide information on all sleep health dimensions. In a correspond- ence addressed to Mohamed et al.’ s study, we discussed two key points for further consideration in future studies: inter- and intraindividual variability in sleep among ME/ CFS [14]. In other words, are sleep disturbances a “one size fits all” ME/CFS? Are they consistent over time? Indeed, people living with ME/CFS form a heterogenous group in terms of disease severity, symptoms range, and intensity, as well as the context in which the disease develops. More importantly, like several other symptoms, sleep disturbances in ME/CFS can vary within the same individual and fluctu- ate over time. Therefore, PSG assessment, over 1 or 2 days may not always capture the full extent of sleep disturbances. Moreover, taking part in a laboratory sleep assessment may not be possible when patients experience increased symp- toms. This could introduce a bias in the evaluation of sleep disturbances within this population [14]. Only a few studies have assessed sleep in people with ME/CFS in their home environment under free-living con- ditions. Available evidence suggests increased TIB and lower estimated SE, together with longer SOL and increased WASO compared to healthy controls [15–18]. However, to our knowledge, no study has examined night-to-night sleep fluctuations or stratified sleep patterns by disease sever- ity. The primary aim of this study was to compare habitual sleep, measured via accelerometry, between individuals with ME/CFS and matched healthy controls. We also evalu- ated sleep across ME/CFS severity subgroups to explore whether greater clinical severity corresponds to more pro- nounced sleep disturbance. Beyond mean sleep duration and quality, we quantified intraindividual variability (IIV), including temporal irregularity in sleep timing and night- to-night fluctuations in SE and related sleep parameters. Because most existing evidence relies on a single-night polysomnography assessment, we further tested whether one night of sleep measurement can reliably represent a ### Page 3 Journal of Clinical Sleep Medicine (2026) 22:77 Page 3 of 13 77patient’s SE. We hypothesized that individuals with ME/ CFS would exhibit greater sleep disruption and increased IIV compared with controls.MethodsStudy design and participantsThis microlongitudinal matched case-control study reports results from the “IN-MOVEMENT” project implemented by the J-AP2S laboratory (University of Toulon, France), in collaboration with the French Association of Chronic Fatigue (www. asso- sfc. org, ASFC, Nice, France). Patients with ME/CFS and their matched healthy controls were recruited through e-mail correspondence delivered by both institutions. Pairs of groups were directly matched 1:1 by age (± 5 years), sex (male or female), and BMI (± 4 kg/m2 ). Eligible par- ticipants provided verbal and written consent and all were adults. Healthy controls were selected based on the absence of both past and present fatiguing conditions or primary sleep disorders, and were considered to be in good overall health. For eligibility, participants with ME/CFS were required to have a formal diagnosis established by a registered physi- cian and to meet the International Consensus Criteria (ICC) [19]. Exclusion criteria were assessed through a structured interview conducted by a trained investigator, focusing on participants’ medical history. The main exclusion criteria were a diagnosed primary psychiatric disorder or a malig- nant disease within the past five years. Additional exclusion criteria included addictions and pregnancy. No formal objec- tive diagnostic sleep assessments (e.g., polysomnography) were performed as part of the study protocol. However, par- ticipants reporting a diagnosed sleep disorder or a medical condition known to substantially affect sleep were excluded. Furthermore, individuals with ME/CFS with severe symp- toms at rest or on a continuous basis according to the Bell Chronic Fatigue Syndrome (CFS) Disability Scale were also excluded from the study, as they were mainly bedridden and thus unable to manage and tolerate the data collection pro- cedure. Ethical clearance for this study was obtained from an Institutional Review Board (IRB00012476-2024-2103–302, which ensured that all assessments adhered strictly to the ethical principles set out in the Declaration of Helsinki.ProcedureAfter enrollment, participants completed an interview to collect sociodemographic information (e.g., age, sex, employment status) and medical history, including prior infections, duration of ME/CFS symptoms, treatments, and comorbidities. Afterward, the follow-up was conducted in the participants’ habitual environments. Each participant received a package by mail containing detailed instruc- tions, a wrist-worn accelerometer (GT3X+, ActiGraph LLC, Pensacola, FL, USA), and a booklet of question- naires. The booklet included a sleep diary to be completed throughout the microlongitudinal follow-up [20], as well as the following questionnaires: the Pittsburgh Sleep Quality Index (PSQI) [21], and the Epworth Sleepiness Scale [22]. Furthermore, participants with ME/CFS were stratified into three groups based on their scores on the Bell CFS Disability Scale: moderate to severe (20–30), moderate (31–49), and mild to moderate (≥ 50), consistent with prior literature on ME/CFS functional classification [11, 23]. Upon completion of the follow-up, participants returned the package, by post to the laboratory.MeasuresSleep monitoring and analysesAccelerometry provides a non-invasive, ecologically valid method for sleep assessment in real-life conditions. How- ever, accelerometry does not directly measure sleep stages or architecture and should therefore be considered a proxy- based method rather than a diagnostic tool. Consequently, sleep parameters derived from accelerometry represent esti- mated indices of sleep-wake patterns and must be interpreted with appropriate caution. Raw triaxial data were collected using ActiGraph GT3X+ devices (ActiGraph LLC, Pen- sacola, FL, USA) worn continuously on the non-dominant wrist for seven consecutive days, excluding water-based activities, at a sampling rate of 60 Hz. Data were processed in R (version 4.0.5; R Core Team, 2021) using the open- source GGIR package (version.3) to estimate sleep param- eters [24]. GGIR operates on raw acceleration data and internally aggregates to short epochs for analysis, applies automatic calibration and non-wear detection, and sup- ports integration of sleep diaries. Sleep-wake classification relied on GGIR’s Heuristic Decision Tree algorithm vali- dated against polysomnography in free-living settings [25]. Daily sleep diaries provided self-reported bed and wake times, which were used as anchors to refine detection of the main nocturnal sleep episode and exclude naps. Sleep diary data were used as temporal anchors to define bedtime and wake-up time windows for accelerometry scoring. Diary- reported lights-off and final awakening times guided the identification of total time in bed (TIB). When diary entries were available, accelerometry data were aligned with the diary reports. Discrepancies between diary and accelerom- etry estimates were uncommon and were resolved through visual inspection of the accelerometry signal (e.g., activity counts) to determine the most plausible TIB interval. When diary data were missing, TIB intervals were identified using ### Page 4 Journal of Clinical Sleep Medicine (2026) 22:7777 Page 4 of 13accelerometry-based algorithms, complemented by visual inspection when needed. Diary completion was requested on a daily basis, and adherence to the study protocol was very high (≈98% of nights). Outputs were visually inspected for plausibility and artifacts (i.e., device non-wear). Extracted sleep estimated metrics included bedtime, wake-up time, TIB, TST, SE, WASO, and SOL.Intraindividual variability (IIV) of sleepIIV of sleep parameters was quantified using three comple- mentary indices: the Coefficient of Variation (CV), the Root Mean Square of Successive Differences (RMSSD), and the Bayesian Variability Model (BVM) [26, 27]. These indices were selected because they capture distinct and complemen- tary dimensions of night-to-night variability. All participants contributed the same number of nights, ensuring that vari- ability estimates were not influenced by differences in data availability. The Coefficient of Variation (CV) reflects the overall magnitude of variability relative to an individual’s mean level. In practical terms, CV indicates how large the fluc- tuations are compared with the typical value of the sleep parameter. Higher CV values therefore represent greater global dispersion across nights. For each participant, CV was calculated as the ratio of the within-person standard deviation to the mean value across all valid nights, expressed as a percentage. The RMSSD captures short-term instability by quantifying the average magnitude of changes between consecutive nights. Unlike CV, which summarizes overall dispersion, RMSSD is sensitive to the night-to-night irreg- ularity of sleep patterns. Higher RMSSD values indicate greater sequential variability (i.e., less stability from one night to the next). Finally, the Bayesian Variability Model (BVM) was used to estimate latent (true) intraindividual variability by modeling nightly values as probabilistic devia- tions around each participant’s mean. This approach pro- vides an estimate of within-person variance while explicitly accounting for measurement uncertainty. Conceptually, the BVM differs from conventional dispersion metrics (e.g., CV, RMSSD) because it separates systematic (latent) vari- ability from residual variability. This distinction yields a more conservative and uncertainty-aware estimate of night- to-night instability. However, residual variability, although often treated as random error, may still reflect meaningful context-dependent or physiological fluctuations. For trans- parency, the mathematical specification of the IIV indices is provided in Table S1.Questionnaires– Pittsburgh Sleep Quality Index (PSQI): Subjective sleep quality was assessed using the PSQI, a validated 19-item questionnaire evaluating sleep over the past month [21]. It yields seven component scores (e.g., sleep latency, duration, disturbances), each rated 0–3. The sum produces a global score from 0 to 21, with scores > 5 indicating poor sleep quality. – Epworth sleepiness scale (ESS): Daytime sleepiness was measured using the ESS, an 8-item self-report assessing general level of daytime sleepiness in com- mon situations [22]. Each item is rated on a 0–3 scale, yielding a total score from 0 to 24; scores > 10 suggest excessive sleepiness. – Bell CFS Disability Scale: The Bell CFS Disability Scale is a self-report tool used to assess functional impairment in individuals with ME/CFS. It ranges from 0 (severely disabled) to 100 (fully functional), with intermediate scores reflecting varying degrees of physical activity, work capacity, and symptom severity [23]. Despite its specificity to people with ME/CFS, this scale is not formally validated through standard psycho- metric procedures (e.g., construct validity, test-retest reliability, factor analysis) in peer-reviewed literature.Statistical analysesThe primary analysis compared people living with ME/ CFS and matched healthy controls. With 38 participants per group and a two-sided α level of 0.05 for the primary analysis, the study provides approximately 80% power to detect standardized mean differences of about d ≈ 0.65, and > 90% power for effects ≥ 0.75. This sample size was therefore considered sufficient to detect moderate-to-large effects in the primary case-control comparisons. Second- ary analyses stratified by disease severity were predefined as exploratory and were not powered for confirmatory inference. For the primary case-control analyses, p-val- ues were used to aid statistical inference together with effect size estimates, whereas secondary analyses were interpreted descriptively, with emphasis on effect sizes and patterns of results rather than on statistical signifi- cance. Because the primary analyses addressed a single overarching research question and involved multiple highly correlated sleep outcomes derived from accelerometry, no formal correction for multiple comparisons was applied in order to limit inflation of type II error. All analyses were performed in RStudio using R version 4.4.1 (R Core Team, 2024). Visualizations were generated with ggplot2, and figure assembly was conducted using ggpubr. Descriptive statistics are reported as mean ± stand- ard deviation unless otherwise specified. Between-group comparisons (ME/CFS vs controls) for continuous vari- ables were examined using independent-samples t-tests, ### Page 5 Journal of Clinical Sleep Medicine (2026) 22:77 Page 5 of 13 77and effect sizes were calculated as Cohen’s d. Compari- sons across ME/CFS severity subgroups were performed using Kruskal-Wallis tests, with Eta 2 reported as the effect size; when significant, Dunn post hoc tests with Holm correction were applied. Categorical variables were analyzed using Chi-square tests or Fisher’s exact tests, with effect sizes expressed as Cramer’s V. The same statistical approach (t-tests for two-group comparisons; Kruskal-Wallis tests for subgroup analyses) was applied to habitual sleep measures and intraindividual variability indices (CV, RMSSD, and BVM). As a sensitivity analy- sis, primary case-control comparisons were re-estimated using linear regression models including Group (ME/CFS vs controls) and employment status (binary: employed vs non-employed) as predictors to account for potential dif- ferences in daytime structure.ResultsSample characteristics and subjective sleepThirty-eight individuals with ME/CFS and matched healthy controls were enrolled in this study. Sociodemographic, clini- cal, and subjective sleep characteristics are summarized in Table 1, and further information on comorbidities and ongo- ing treatments is provided in Table S2. Compared with healthy controls, people with ME/CFS reported poorer subjective sleep quality, with large effect sizes across all PSQI domains except sleep duration. Notably, 95% of participants with ME/ CFS exceeded the clinical PSQI cutoff (> 5), indicating per- vasive and clinically meaningful sleep disturbances compared to controls (p < 0.001). Daytime sleepiness, assessed by ESS, was also significantly elevated in the ME/CFS group relative to their healthy counterparts (p < 0.001) with 47% showing excessive daytime sleepiness. Subjective sleep disturbances were consistently observed across all disease severity sub- groups. However, the proportion of participants with excessive sleepiness (ESS > 10) did differ significantly across severity levels (p = 0.014, large effect size = 0.48).Comparison of habitual sleep between ME/CFS Patients and Healthy Controls (including severity subgroups)As shown in Table 2, compared with healthy controls, individuals with ME/CFS did not differ significantly in bedtime (p = 0.053, ES = 0.45 [Medium effect]), later morning wake times (p < 0.05, ES = 0.46 [Medium effect]), resulting in significantly longer TIB (p < 0.001). How- ever, ME/CFS showed reduced accelerometry estimated SE (p < 0.001), marked by a nearly doubled sleep onset latency (p < 0.001) and increased wake after sleep onset (p < 0.001). Despite spending nearly an extra hour in bed, TST did not differ between the two groups. Sleep distur- bances appeared broadly similar across severity subgroups, with negligible effect sizes observed between strata.Intraindividual variability in sleep between ME/CFS and healthy controls (including severity subgroups)IIV of sleep including temporal instability of sleep timing and night-to-night fluctuations in sleep metrics are pre- sented in Table 3.Coefficient of variation (CV)Compared with healthy controls, individuals with ME/CFS showed significantly higher variability in estimated sleep metrics including duration, and continuity, as indicated by increased CV for TST (p = 0.010, ES = 0.61 [Medium]), SE (p < 0.001, ES = 3.64 [Large]), and WASO (p < 0.001, ES = 0.94 [Large]). In contrast, variability in bedtime and TIB was lower in ME/CFS (p = 0.014, ES = 0.58 [Medium]; p = 0.048, ES = 0.46 [Medium], respectively), while no group differences were observed for wake-up time or SOL (all p > 0.05, ES = 0.08–0.13). Across ME/ CFS severity subgroups, exploratory comparisons did not reveal clear differences in CV parameters, with observed effect sizes remaining negligible to small.Root mean square of successive differences (RMSSD)ME/CFS patients displayed greater night-to-night variability in both sleep initiation and continuity, reflected by higher RMSSD values for TST (p = 0.031, ES = 0.50 [Medium]), SE (p < 0.001, ES = 2.67 [Large]), SOL (p < 0.001, ES = 0.84 [Large]), and WASO (p < 0.001, ES = 1.30 [Large]). Wake- up time instability was also elevated in ME/CFS compared to healthy controls (p = 0.004, ES = 0.69 [Large]), whereas bed- time instability was lower (p = 0.045, ES = 0.47 [Medium]), and TIB showed no substantial group differences (ME/CFS vs. controls; p > 0.05). Similar to CV results, no RMSSD clear differences emerged between ME/CFS severity sub- groups (ES = negligible–small).Bayesian variability model (BVM)BVM estimates revealed no significant differences between groups for wake-up time, TIB, TST, SE, SOL, or WASO (allp > 0.33, ES = negligible–small). Only bedtime variability remained lower in ME/CFS compared to controls (p = 0.038, ### Page 6 Journal of Clinical Sleep Medicine (2026) 22:7777 Page 6 of 13Table 1 Demographics, clinical, and subjective sleep related characteristics among people with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) including severity subgroups compared to healthy controls Controls ME/CFS ME/CFS [Mild to Moder- ate] ME/CFS [Moderate] ME/CFS [Moderate to Severe] (N = 38) (N = 38) p Effect Size [Interpretation] (N = 5) (N = 20) (N = 13) p Effect Size [Interpretation]Demographic dataAge (years) 48 ± 11.88 51 ± 10.13 0.240 0.27 [Small] 57 ± 10.84 53.7 ± 10.09 44.54 ± 6.60 0.003 0.28 [Large] BMI (kg/m 2 ) 23.68 ± 4.01 23.03 ± 4.44 0.500 0.16 [Small] 22.60 ± 5.46 24 ± 4.94 21.69 ± 2.93 0.508 0.02 [Negligible] Sex 1.000 0 [Negligible] 0.346 0.26 [Medium] Female (N, %) 30, 78.95% 30, 78.95% 5, 100% 14, 70% 11, 84.62% Male (N, %) 8, 21.95% 8, 21.95% 0, 0% 6, 30% 2, 15.38% Employment status < 0.001 0.71 [Large] 0.019 0.48 [Large] Employed (N,%) 27, 71.05% 6, 15.79% 3, 60% 1, 5% 2, 15.38% Unemployed (N, %) 5, 13.16% 3, 7.89% 0, 0% 3, 15% 0, 0% Retired (N, %) 2, 5.26% 5, 13.16% 1, 20% 4, 20% 0, 0% Unable to work (N, %) 0, 0% 22, 57.89% 1, 20% 10, 50% 11, 84.62% Homemaker (N, %) 4, 10.53% 2, 5.26% 0, 0% 2, 10% 0, 0%Clinical dataDuration of illness (months) N/A 129.90 ± 123.99 134.4 ± 81.56 138 ± 139.70 87 ± 73.28 0.377 < 0.01 [Negligi- ble] Diagnostic delay (months) N/A 34.71 ± 46.71 27.8 ± 31.63 49 ± 60.36 20.46 ± 21.60 0.239 0.02 [Negligible] Bell CFS Disability Scale N/A 40.54 ± 10.53 58.75 ± 2.50 44.25 ± 5.68 29.23 ± 2.77 < 0.001 0.84 [Large]Subjective sleep-related dataPSQI global score Subjective sleep quality 0.45 ± 0.50 1.63 ± 0.71 < 0.001 1.92 [Large] 1.8 ± 0.84 1.5 ± 0.69 1.77 ± 0.73 0.681 0.04 [Negligible] Sleep latency 1.21 ± 0.58 1.76 ± 0.94 0.003 0.71 [Large] 1.6 ± 1.14 1.85 ± 0.99 1.69 ± 0.85 0.846 0.05 [Negligible] Sleep duration 0.42 ± 0.50 0.45 ± 0.76 0.859 0.04 [Negligible] 0.80 ± 1.30 0.45 ± 0.69 0.31 ± 0.63 0.667 0.03 [Negligible] Sleep efficiency 0.47 ± 0.51 1.35 ± 1.34 < 0.001 0.87 [Large] 1.6 ± 1.52 1.45 ± 1.32 1.08 ± 1.38 0.622 0.03 [Negligible] Sleep disturbance 0.45 ± 0.50 1.76 ± 0.54 < 0.001 2.51 [Large] 1.6 ± 0.55 1.75 ± 0.55 1.85 ± 0.55 0.694 0.04 [Negligible] Need for sleep inducers 0.03 ± 1.16 0.87 ± 1.32 < 0.001 0.90 [Large] 1.6 ± 1.52 0.7 ± 1.26 0.85 ± 1.34 0.383 0.002 [Negligible] ### Page 7 Journal of Clinical Sleep Medicine (2026) 22:77 Page 7 of 13 77ES = 0.48 [Medium]). No between-severity differences were obvious for any BVM metric (all ES = negligible).Comparison of intraindividual variability metrics resultsAnalyses of intraindividual variability revealed distinct pat- terns depending on the variability metric applied. Both CV and RMSSD consistently indicated greater night-to-night fluctuation in several sleep parameters among individuals with ME/CFS compared with controls. These indices, which capture overall dispersion across nights (CV) and short-term successive fluctuations (RMSSD), suggest that sleep in ME/ CFS is characterized by higher variability. In contrast, BVM, which isolates stable intraindividual variance while account- ing for measurement uncertainty, did not identify correspond- ing group differences, indicating the absence of a structured or systematic pattern of variability. This suggests that the insta- bility detected by conventional indices may reflect irregular or context-dependent fluctuations rather than consistent within- person variance. Notably, BVM-derived variability in bedtime remained significantly lower in the ME/CFS group compared with controls (p = 0.038), consistent with reduced variability in sleep timing previously reported using CV and RMSSD.Sensitivity analysesIn sensitivity analyses adjusting for employment status, the primary case-control differences in estimated SE, SOL, WASO, and key variability indices remained similar. Employment status was not independently associated with the sleep outcomes (Table S3).Estimated sleep efficiency variability in ME/CFS challenges single‑night evaluationFigure 1a shows that estimated SE in ME/CFS exhibits pro- nounced night-to-night variability, with frequent alternations between low- and high-efficiency nights, whereas controls dis- play much more stable patterns. As illustrated in Fig. 1b, this variability translates into a significantly different distribution of estimated SE categories between groups (χ2 (2) = 128.5,p < 0.001), with a moderate-to-large effect size (Cramer’sV = 0.49). Consequently, a single-night assessment may fail to capture low-efficiency nights or may overrepresent high- efficiency nights, thereby underestimating the overall burden and variability of sleep disturbances in ME/CFS.DiscussionThe present study provides a refined characterization of habitual sleep disturbances in ME/CFS under real-life condi- tions by moving beyond mean sleep parameters to explicitlyTable 1 (continued) Controls ME/CFS ME/CFS [Mild to Moder- ate] ME/CFS [Moderate] ME/CFS [Moderate to Severe] (N = 38) (N = 38) p Effect Size [Interpretation] (N = 5) (N = 20) (N = 13) p Effect Size [Interpretation] Daytime dysfunction 0.08 ± 0.27 1.55 ± 0.92 < 0.001 2.17 [Large] 1.4 ± 0.89 1.45 ± 0.89 1.77 ± 1.01 0.462 0.01 [Negligible] Total 3.11 ± 0.39 9.37 ± 3.17 < 0.001 2.78 [Large] 10.40 ± 4.16 9.15 ± 3.34 9.31 ± 2.63 0.771 0.04 [Negligible] Score > 5 (N, %) 0, 0% 35, 94.59% < 0.001 0.92 [Large] 5, 100% 17, 89.47% 13, 100% 0.629 0.23 [Medium] ESS score Score > 10 (N, %) 0, 0% 18, 47.38% < 0.001 0.53 [Large] 4, 80% 5, 25% 9, 69.23% 0.014 0.48 [Large] Total 4.45 ± 1.18 10.26 ± 6.52 < 0.001 1.24 [Large] 10.60 ± 6.54 8.50 ± 6.08 12.84 ± 6.77 0.140 0.06 [Small] Data are presented as mean ± standard deviation or as number of participants and percentages. Duration of illness refers to the time since symptom onset, and diagnostic delay to the time until formal diagnosis. Employment categories represent self-reported current statusBMI, body mass index; CFS, chronic fatigue syndrome; ESS, Epworth Sleepiness Scale; ME/CFS, myalgic encephalomyelitis/chronic fatigue syndrome; N, number of participants; N/A, not applicable; PSQI, Pittsburgh Sleep Quality Index; SD, standard deviation. Bold values indicate statistically significant differences between groups (p<0.05) ### Page 8 Journal of Clinical Sleep Medicine (2026) 22:7777 Page 8 of 13examine night-to-night fluctuation. While individuals with ME/CFS spent more time in bed and exhibited lower esti- mated SE than matched controls, total sleep time did not differ between groups. This dissociation reinforces prior evi- dence that prolonged sleep opportunity in ME/CFS does not translate into restorative sleep, and suggests that the primary disturbance may lie in the capacity to initiate and maintain sleep rather than sleep duration [15–18]. Importantly, the current findings extend this literature by demonstrating that impaired estimated SE in ME/CFS is not only reduced on average but also highly unstable across nights, even in the presence of relatively regular sleep timing compared to healthy controls. Sleep normally exhibits some degree of night-to-night fluctuation in response to environmental, psychosocial, and physiological influences [28]. However, the magnitude of variability observed in ME/CFS exceeded that reported in healthy populations and was comparable to levels docu- mented in clinical sleep disorders such as insomnia [29–31]. For instance, using accelerometry, Veeramachaneni et al. reported a within-person standard deviation in estimated SE of 5.83% in young adults with insomnia, corresponding to a CV of approximately 7.0% [29]. Buysse et al. observed a slightly higher SE variability in older adults with chronic insomnia (SD = 7.0%; CV ≈ 8.6%) [30], and Chan et al. reported a within-person standard deviation of 6.7%, cor- responding to an estimated CV of approximately 8–8.5% [31]. Notably, this heightened estimated SE variability occurred despite preserved or even reduced variability in bedtime, contrasting with insomnia populations in which irregular sleep timing often accompanies poor sleep continu- ity [29–31]. This dissociation challenges maladaptive sleep behaviors and may instead point toward instability within physiological systems governing sleep maintenance. Con- sistent with this interpretation, prior work has shown that excessive night-to-night variability in sleep continuity is associated with inflammatory processes, autonomic dysregu- lation, and altered homeostatic sleep pressure, independent of mean sleep duration or efficiency [32]. In ME/CFS, such mechanisms are highly plausible [33, 34] A key contribution of the present study is the simultane- ous application of complementary variability metrics. Con- ventional dispersion indices (CV and RMSSD) consistently indicated greater night-to-night fluctuations in estimated SE and related parameters in ME/CFS. In contrast, the Bayes- ian variability model, which isolates stable latent variabil- ity while accounting for measurement uncertainty, did not reveal corresponding group differences. This divergence may suggest that the observed variability may primarily reflect context-dependent fluctuations rather than a fixed, trait-like increase in underlying variability. Importantly, the residual component captured by CV and RMSSD may in fact encode physiologically meaningful responses to daily perturbations. In ME/CFS, such perturbations could include variations inTable 2 Habitual sleep among people with myalgic encephalomyelitis/chronic fatigue syndrome (including severity subgroups) compared to healthy controls Data are presented as mean ± standard deviation. Sleep parameters were obtained from one week of continuous wrist-worn accelerometry. Bed- time and wake-up time represent the average timing of nocturnal sleep episodes. Total time in bed refers to the duration between bedtime and final wake-up time, while total sleep time represents the estimated amount of sleep within this period. Sleep efficiency is calculated as the ratio of total sleep time to total time in bed, expressed as a percentage. Sleep onset latency indicates the time required to transition from wakefulness to sleep, and wake after sleep onset reflects the total duration of awakenings occurring after sleep onsetME/CFS, myalgic encephalomyelitis/chronic fatigue syndrome; SD, standard deviation; SE, Sleep Efficiency; SOL, Sleep Onset Latency; TIB, Time in Bed; TST, Total Sleep Time; WASO, Wake After Sleep Onset. Bold values indicate statistically significant differences between groups (p<0.05) Controls ME/CFS ME/CFS [Mild to Moderate] ME/CFS [Moderate] ME/CFS [Moderate to Severe] (N = 38) (N = 38) p Effect Size [Interpreta- tion] (N = 5) (N = 20) (N = 13) p Effect Size [Interpreta- tion] Bedtime (hh:mm ± hh:min) 23:06 ± 00:25 22:44 ± 01:05 0.053 0.45 [Medium] 22:01 ± 00:34 22:43 ± 00:57 23:02 ± 01:20 0.125 0.06 [Small] Wake-up time (hh:mm ± hh:min) 07:18 ± 00:23 07:44 ± 01:15 0.048 0.46 [Medium] 07:15 ± 00:28 07:38 ± 01:29 08:04 ± 00:59 0.125 0.06 [Small] TIB (hh:mm ± min) 08:11 ± 00:35 08:59 ± 01:15 < 0.001 0.80 [Large] 09:13 ± 00:47 08:55 ± 01:24 09:01 ± 01:16 0.871 0.05 [Small] TIB (min) 491.58 ± 35.99 539.32 ± 75.77 < 0.001 0.80 [Large] 553.4 ± 46.84 534.9 ± 83.69 540.69 ± 75.91 0.871 0.05 [Small] TST (hh:mm ± min) 06:48 ± 00:31 06:55 ± 00:53 0.491 0.16 [Small] 07:00 ± 00:49 06:52 ± 00:58 06:57 ± 00:49 0.950 0.05 [Small] TST (min) 407.68 ± 31.08 414.58 ± 52.90 0.491 0.16 [Small] 420 ± 48.63 411.90 ± 58.31 416.62 ± 49.27 0.950 0.05 [Small] SE (%) 82.5 ± 1.89 76.71 ± 3.55 < 0.001 2.04 [Large] 75.40 ± 2.70 76.85 ± 3.54 77 ± 3.96 0.749 0.04 [Small] SOL (min) 19.05 ± 7.01 33.21 ± 12.88 < 0.001 1.37 [Large] 29.20 ± 3.42 31.30 ± 12.68 37.69 ± 7.01 0.381 0.001 [Neg- ligible] WASO (min) 63.82 ± 13.75 90.79 ± 26.93 < 0.001 1.26 [Large] 103.60 ± 11.57 91.20 ± 26.51 85.23 ± 31.25 0.171 0.04 [Small] ### Page 9 Journal of Clinical Sleep Medicine (2026) 22:77 Page 9 of 13 77symptom burden, physical or cognitive exertion, and post- exertional malaise (PEM). The finding of reduced bedtime variability alongside unstable SE further supports this inter- pretation, suggesting that individuals with ME/CFS may actively maintain regular sleep schedules as an adaptive pac- ing strategy, yet remain unable to achieve consistent sleep continuity due to fluctuating internal constraints. We there- fore propose that pronounced night-to-night sleep variability in ME/CFS may partly reflect the characteristic “push-and- crash” dynamic of the illness. Periods of increased activity may precipitate delayed physiological deterioration, mani- festing as PEM and subsequent sleep fragmentation [35, 36]. Analogous patterns have been observed during acute infec- tion, where sleep duration may increase while continuity deteriorates as part of an inflammation-driven response [37]. In ME/CFS, even minor daily activities could trigger such responses, leading to unpredictable alternations between relatively preserved and markedly impaired nights of sleep. However, sleep under free-living conditions is also influ- enced by daytime structure and environmental context [32]. Although employment status differed markedly between groups, sensitivity analyses adjusting for employment status did not materially alter the primary findings, suggesting that the observed sleep disturbances and variability in ME/CFSTable 3 Intraindividual variability of sleep among people with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) including sever- ity subgroups compared to healthy controls Data are presented as mean ± standard deviation. Intraindividual variability (IIV) of sleep was calculated across one week of accelerometry recordings. CV expresses relative variability as a percentage. RMSSD quantifies short-term night-to-night fluctuations and is expressed in the same units as the corresponding parameter. BVM provides a standardized, unitless estimate of true within-person variance after accounting for measurement error and uncertaintya.u., arbitrary units; BVM, Bayesian Variability Model; CV, coefficient of variation; ME/CFS, myalgic encephalomyelitis/chronic fatigue syn- drome; RMSSD, root mean square of successive differences; SD, standard deviation; SE, sleep efficiency; SOL, sleep onset latency; TIB, total time in bed; TST, total sleep time; WASO, wake after sleep onset. Bold values indicate statistically significant differences between groups (p<0.05) Controls ME/CFS ME/CFS [Mild to Moderate] ME/CFS [Moderate] ME/CFS [Moderate to Severe] Mean ± SD (N = 38) Mean ± SD (N = 38)p Effect Size [Interpretation] Mean ± SD (N = 5) Mean ± SD (N = 20) Mean ± SD (N = 13)p Effect Size [Interpretation]CVBedtime (%) 2.97 ± 0.72 2.32 ± 1.44 0.014 0.58 [Medium] 2 ± 0.71 2.5 ± 1.67 2.15 ± 1.28 0.823 0.05 [Small] Wake-up time (%) 9.05 ± 1.75 9.58 ± 5.45 0.573 0.13 [Small] 8.8 ± 4.55 9.45 ± 5.33 10.08 ± 6.26 0.927 0.05 [Small] TIB (%) 11.13 ± 3.73 9.13 ± 4.87 0.048 0.46 [Medium] 7 ± 1.58 9.65 ± 5.19 9.15 ± 3.73 0.513 0.02 [Negligible] TST (%) 11.29 ± 3.73 13.87 ± 4.70 0.010 0.61 [Medium] 12.40 ± 3.58 13.6 ± 3.38 14.85 ± 6.61 0.559 0.02 [Negligible] SE (%) 2.53 ± 1.03 10.13 ± 2.77 < 0.001 3.64 [Large] 10.6 ± 1.52 9.7 ± 2.70 10.62 ± 3.28 0.405 0.01 [Negligible] SOL (%) 57.29 ± 25.52 59.18 ± 24.61 0.743 0.08 [Negligible] 53.80 ± 13.61 57.80 ± 21.19 63.38 ± 32.56 0.961 0.05 [Small] WASO (%) 29.24 ± 14.47 43.92 ± 16.67 < 0.001 0.94 [Large] 38.20 ± 2.49 44.45 ± 16.50 45.31 ± 20.18 0.481 0.02 [Negligible]RMSSDBedtime (hh:mm ± hh:min) 01:00 ± 00:24 00:48 ± 00:26 0.045 0.47 [Medium] 00:32 ± 00:10 00:52 ± 00:30 00:49 ± 00:23 0.319 0.01 [Negligible] Wake-up time (hh:mm ± hh:min) 00:41 ± 00:07 01:01 ± 00:40 0.004 0.69 [Large] 00:58 ± 00:28 00:56 ± 00:35 01:10 ± 00:52 0.815 0.05 [Small] TIB (min) 76.74 ± 26.41 74.71 ± 43.53 0.807 0.06 [Negligible] 59.80 ± 17.81 74 ± 37.21 81.54 ± 53.31 0.673 0.03 [Negligible] TST (min) 65.24 ± 23.36 80.05 ± 34.35 0.031 0.50 [Medium] 80.40 ± 31.67 76.25 ± 26.97 85.77 ± 45.76 0.926 0.05 [Small] SE (%) 2.66 ± 1.21 10.32 ± 3.87 < 0.001 2.67 [Large] 11.60 ± 2.30 9.65 ± 3.98 10.85 ± 4.20 0.444 0.01 [Negligible] SOL (min) 15.26 ± 9.93 27.39 ± 17.75 < 0.001 0.84 [Large] 21.40 ± 5.77 24.85 ± 17.02 33.62 ± 20.82 0.486 0.02 [Negligible] WASO (min) 26.29 ± 12.08 49.03 ± 21.64 < 0.001 1.30 [Large] 56.20 ± 19.58 48.35 ± 22.11 47.31 ± 22.72 0.702 0.04 [Negligible]BVMBedtime (a.u.) 64.47 ± 9.48 60.16 ± 8.31 0.038 0.48 [Medium] 58 ± 9.82 60 ± 8.18 61.23 ± 8.47 0.677 0.03 [Negligible] Wake-up time (a.u.) 58.66 ± 29.33 60 ± 31.18 0.847 0.04 [Negligible] 77.20 ± 3.03 57.15 ± 29.80 57.77 ± 32.94 0.361 0.001 [Negligible] TIB (a.u.) 80.61 ± 16.17 77.03 ± 15.62 0.330 0.23 [Small] 80.61 ± 19.59 76.27 ± 12.99 78.44 ± 18.74 0.897 0.05 [Small] TST (a.u.) 64.92 ± 9.55 63.71 ± 9.16 0.574 0.13 [Small] 66.20 ± 9.96 62.95 ± 7.87 65.44 ± 10.97 0.703 0.04 [Negligible] SE (a.u.) 5.68 ± 2.47 5.66 ± 2.54 0.964 0.01 [Negligible] 5.52 ± 2.26 6.27 ± 2.85 6.14 ± 2.35 0.946 0.05 [Small] SOL (a.u.) 19.05 ± 7.59 17.50 ± 8.72 0.410 0.19 [Small] 19.40 ± 16.68 18.70 ± 8.52 14.92 ± 3.95 0.563 0.02 [Negligible] WASO (a.u.) 36.05 ± 10.03 37.84 ± 10.04 0.440 0.18 [Small] 35 ± 9.27 38.35 ± 11.54 38.15 ± 8.20 0.781 0.04 [Negligible] ### Page 10 Journal of Clinical Sleep Medicine (2026) 22:7777 Page 10 of 13are unlikely to be solely explained by differences in daytime occupational structure. In addition, circadian disruption has been proposed as a contributing mechanism in ME/CFS, potentially involving altered coordination by the suprachiasmatic nucleus and downstream desynchronization of sleep-wake, autonomic, and metabolic processes [38]. Genetic and transcriptomic studies reported markedly elevated expression of the clock gene NPAS2 in ME/CFS patients, although these investi- gations remain underpowered and lack repeated rhythmic sampling [39]. Studies using accelerometry consistently describe reduced daytime activity, lower rhythm amplitude, and irregular rest-activity patterns. However, interpreta- tion remains limited by masking effects and the absence of constant routine or forced desynchrony protocols capable of isolating endogenous circadian signals [40]. In addi- tion, lifestyle constraints frequently observed in ME/CFS, including lower employment rates and increased time spent indoors, may reduce exposure to light, a primary circadian zeitgeber, thereby potentially exacerbating circadian mis- alignment. Physiological markers show mixed findings, with largely preserved mean temperature and melatonin timing but altered coupling between dim light melatonin onset (DLMO) and temperature rhythms, increased vari- ability, and atypical distal skin temperature profiles [16, 40] Alongside this circadian dysregulation, the variability of sleep patterns becomes particularly relevant. In fact, large- scale studies suggest that irregular sleep patterns, rather than sleep duration alone, are more strongly associated with adverse cardiovascular outcomes and circadian disruption [41, 42]. The present findings are particularly noteworthy, as they demonstrate marked night-to-night sleep variability in ME/CFS despite relatively preserved bedtime regular- ity, pointing to variability of sleep continuity as a distinct feature of the disorder. Recent data further indicate an ele- vated cardiovascular risk in ME/CFS, potentially driven by chronic low-grade inflammation, autonomic dysfunction, and endothelial impairment [43]. Whether sleep irregularity contributes to this increased vulnerability, independently or in interaction with these mechanisms, remains to be deter- mined and warrants investigation in future longitudinal stud- ies [44]. Especially in light of a recent study linking lipid profile alterations in ME/CFS to both circadian rhythm and sleep disturbances, with endothelial dysfunction proposed as an underlying mechanism [45]. Exploratory analyses did not reveal clear differences in estimated SE or variability across ME/CFS severity sub- groups. These null findings should be interpreted cautiously. The severity-stratified analyses were underpowered and lim- ited by a restricted severity range, as individuals with severe symptoms at rest were excluded. It is also plausible that indi- viduals with more severe disease engage in stricter activ- ity limitation to avoid symptom exacerbation, potentially stabilizing sleep patterns despite greater underlying pathol- ogy. Future studies incorporating larger samples across the full severity continuum, alongside concurrent assessment of daily symptoms and activity levels, are needed to clarify how disease severity modulates sleep variability.Fig. 1 Sleep efficiency patterns and night-category distributions in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and controls. a Individual heatmaps showing sleep efficiency across seven consecutive nights in participants with ME/CFS (top panel) and healthy controls (bottom panel). Each cell represents a single night per participant, with brighter colors indicating lower sleep effi- ciency and darker colors reflecting higher efficiency. b Proportion of nights falling into three sleep efficiency categories (< 74%, 74–85%, > 85%) for each group. SE categories were defined according to the National Sleep Foundation’s adult sleep quality recommendations, with SE > 85% considered appropriate, 74–85% uncertain, and < 74% inappropriate [44]. CTL, controls; ME/CFS, myalgic encephalomyeli- tis/chronic fatigue syndrome; SE, sleep efficiency ### Page 11 Journal of Clinical Sleep Medicine (2026) 22:77 Page 11 of 13 77Several limitations should be acknowledged. The sam- ple size in subgroup analyses was relatively small, which may have limited our ability to detect more subtle effects. In addition, the exclusion of sleep disorders relied solely on interview-based medical history rather than systematic poly- somnography. Consequently, the presence of undiagnosed sleep disorders cannot be entirely ruled out. The relatively short monitoring period also restricts the ability to character- ize longer-term patterns, emphasizing the need for extended longitudinal assessments. Although accelerometry provides ecologically valid data, it does not capture sleep architecture or autonomic activity; combining multi-night ambulatory EEG with variability metrics would provide deeper mecha- nistic insight. Furthermore, daily symptom burden, activ- ity levels, and PEM episodes were not modeled alongside variability, limiting the ability to directly link physiological triggers with sleep fluctuations. Clinically, the present findings highlight that the frequent alternation between low- and relatively preserved estimated efficiency nights implies that the specific night selected for clinical evaluation may strongly influence diagnostic impressions and subsequent treatment decisions. Accord- ingly, extended multi-night assessment, ideally under ambulatory conditions (accelerometry, sleep diaries), may provide a more representative evaluation of sleep in this population. In addition, the coexistence of relatively regu- lar sleep timing with increased estimated SE variability suggests that sleep fragmentation in ME/CFS may not pri- marily arise from irregular sleep scheduling, thereby dis- tinguishing it from behavioral insomnia in which timing instability plays a central role. However, these observations also indicate that therapeutic approaches focusing solely on sleep duration or sleep opportunity may be insufficient, and that strategies aimed at reducing night-to-night fluctuation may warrant consideration in the clinical management of this population.ConclusionThis study demonstrates that people living with ME/CFS exhibited impaired sleep initiation and continuity, accompa- nied by pronounced night-to-night variability despite more stable sleep scheduling than controls. By integrating conven- tional sleep parameters with complementary variability met- rics, we show that sleep in ME/CFS is highly variable. These findings underscore the importance of viewing variability as a central feature of ME/CFS-related sleep disturbances and point to the need to address the underlying physiologi- cal instability that may disrupt sleep-wake regulation in this condition.AQSupplementary Information The online version contains supplemen- tary material available at https://doi.org/10.1007/s44470-026-00079-7.Acknowledgements The authors would like to thank the participants for their time and commitment. We also extend our thanks to Robert Schenk from the French Association of Chronic Fatigue (ASFC) for his availability and support.Author contribution MS contributed to data curation, investigation, and writing of the original draft. IF contributed to study conceptualiza- tion and methodology, and was responsible for patient recruitment and feasibility assessment. GCDS contributed to methodology, formal data analysis, and data visualization. CC contributed to methodology and formal analysis. LF and MG contributed to the critical revision of the manuscript. OS conceptualized and designed the study, supervised the project, contributed to data visualization, and led the writing, review, and editing of the manuscript. All authors reviewed and approved the final version of the manuscript.Funding Open access funding provided by Université de Toulon. This work was supported by internal funding from the J-AP2S Laboratory.Data Availability The datasets supporting this analysis will be made available from the corresponding author upon reasonable request.DeclarationsEthics approval and consent to participate The study was approved by an Institutional Review Board (IRB00012476-2024-21-03-302). Written informed consent was obtained from all participants prior to inclusion in the study.Consent for publication Not applicable.Competing interests All authors have read and agreed to the published version of the manuscript. The authors declare no competing interests.Open Access This article is licensed under a Creative Commons Attri- bution 4.0 International License, which permits use, sharing, adapta- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.References1. Deumer U-S, Varesi A, Floris V, Savioli G, Mantovani E, López- Carrasco P, et al. Myalgic encephalomyelitis/chronic fatigue syn- drome (ME/CFS): an overview. J Clin Med. 2021;10:4786. https://doi.org/10.3390/jcm10204786.2. Lim E-J, Son C-G. Review of case definitions for myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). J Transl Med BioMed Central. 2020;18:1–10. https:// doi. org/ 10. 1186/s12967-020-02455-0.3. Cortes Rivera M, Mastronardi C, Silva-Aldana CT, Arcos-Bur- gos M, Lidbury BA. Myalgic encephalomyelitis/chronic fatigue ### Page 12 Journal of Clinical Sleep Medicine (2026) 22:7777 Page 12 of 13syndrome: a comprehensive review. Diagnostics. 2019;9:91. https://doi.org/10.3390/diagnostics9030091.4. Klimas NG, Koneru AO, Springer. Chronic fatigue syndrome: inflammation, immune function, and neuroendocrine interactions. Curr Rheumatol Rep. 2007;9:482–7. https:// doi. org/ 10. 1007/s11926-007-0078-y.5. Missailidis D, Annesley SJ, Fisher PR. Pathological mechanisms under- lying myalgic encephalomyelitis/chronic fatigue syndrome. Diagnos- tics. 2019;9:80. https://doi.org/10.3390/diagnostics9030080.6. Vardaman M, Gilmour S. Letter: time to correct the record on the global burden of myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). J Transl Med. 2025;23:331. https://doi.org/10.1186/s12967-025-06281-0.7. Xiong R, Aiken E, Caldwell R, Vernon SD, Kozhaya L, Gunter C, et al. AI-driven multi-omics modeling of myalgic encephalo- myelitis/chronic fatigue syndrome. Nat Med. 2025;31:2991–3001. https://doi.org/10.1038/s41591-025-03788-3.8. Rasa S, Nora-Krukle Z, Henning N, Eliassen E, Shikova E, Harrer T, et al. Chronic viral infections in myalgic encephalomyelitis/ chronic fatigue syndrome (ME/CFS). J Transl Med. 2018;16:1– 25. https://doi.org/10.1186/s12967-018-1644-y.9. Davis HE, McCorkell L, Vogel JM, Topol EJ. Long COVID: major findings, mechanisms and recommendations. Nat Rev Microbiol. 2023;21:133–46. https://doi.org/10.1038/s41579-022-00846-2.10. Kedor C, Freitag H, Meyer-Arndt L, Wittke K, Hanitsch LG, Zoller T, et al. A prospective observational study of post- COVID-19 chronic fatigue syndrome following the first pan- demic wave in Germany and biomarkers associated with symptom severity. Nat Commun. 2022;13:5104. https://doi.org/10.1038/s41467-022-32507-6.11. Bateman L, Bested AC, Bonilla HF, Chheda BV, Chu L, Cur- tin JM, et al. Myalgic Encephalomyelitis/Chronic Fatigue Syn- drome: essentials of diagnosis and management. Mayo Clin Proc. 2021;96:2861–78. https://doi.org/10.1016/j.mayocp.2021.07.004.12. Clayton EW, American Medical Association. Beyond myalgic encephalomyelitis/chronic fatigue syndrome: an IOM report on redefining an illness. JAMA. 2015;313:1101–2. https://doi.org/10.1001/jama.2015.1346.13. Mohamed AZ, Andersen T, Radovic S, Del Fante P, Kwiatek R, Calhoun V, et al. Objective sleep measures in chronic fatigue syn- drome patients: A systematic review and meta-analysis. Sleep Med Rev. 2023;69:101771. https://doi.org/10.1016/j.smrv.2023.101771.14. Saidi O, Fantini M-L, Gruet M, Duché P. Commentary on Mohamed et al.: inter- and intra-sleep variability deserve further attention in myalgic encephalomyelitis/chronic fatigue syndrome. Sleep Med Rev. 2023;70:101812. https://doi.org/10.1016/j.smrv.2023.101812.15. Aerenhouts D, Ickmans K, Clarys P, Zinzen E, Meersdom G, Lambrecht L, et al. Sleep characteristics, exercise capacity and physical activity in patients with chronic fatigue syndrome. Disa- bil Rehabil. 2015;37:2044–50. https://doi.org/10.3109/09638288.2014.993093.16. Cambras T, Castro-Marrero J, Zaragoza MC, Díez-Noguera A, Alegre J. Circadian rhythm abnormalities and autonomic dysfunc- tion in patients with Chronic Fatigue Syndrome/Myalgic Encepha- lomyelitis. PLoS One. 2018;13:e0198106. https://doi.org/10.1371/journal.pone.0198106.17. Libman E, Creti L, Baltzan M, Rizzo D, Fichten CS, Bailes S. Sleep apnea and psychological functioning in chronic fatigue syndrome. J Health Psychol. 2009;14:1251–67. https://doi.org/10.1177/1359105309344895.18. Rahman K, Burton A, Galbraith S, Lloyd A, Vollmer-Conna U. Sleep-wake behavior in chronic fatigue syndrome. Sleep. 2011;34:671–8. https://doi.org/10.1093/sleep/34.5.671.19. Carruthers BM, van de Sande MI, De Meirleir KL, Klimas NG, Broderick G, Mitchell T, et al. Myalgic encephalomyelitis: inter- national consensus criteria. J Intern Med. 2011;270:327–38. https://doi.org/10.1111/j.1365-2796.2011.02428.x.20. Carney CE, Buysse DJ, Ancoli-Israel S, Edinger JD, Krystal AD, Lichstein KL, et al. The consensus sleep diary: standardizing pro- spective sleep self-monitoring. Sleep. 2012;35:287–302. https://doi.org/10.5665/sleep.1642.21. Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh sleep quality index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28:193–213. https://doi.org/10.1016/0165-1781(89)90047-4.22. Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep. 1991;14:540–5. https://doi.org/10.1093/sleep/14.6.540.23. Bell DS. The doctor’s guide to chronic fatigue syndrome: under- standing, treating, and living with CFIDS. Addison-Wesley Pub- lishing Company; 1994.24. Migueles JH, Rowlands AV, Huber F, Sabia S, van Hees VT, et al. GGIR: a research community–driven open source R package for generating physical activity and sleep outcomes from multi-day raw accelerometer data. J Meas Phys Behav. 2019;2:188–96. https://doi.org/10.1123/jmpb.2018-0063.25. Hees VT, Sabia S, Anderson KN, Denton SJ, Oliver J, Catt M, et al. A novel, open access method to assess sleep duration using a wrist-worn accelerometer. PLoS ONE. 2015;10:e0142533. https://doi.org/10.1371/journal.pone.0142533.26. Heaney M, Healy O, Wyse J, White A. Bayesian latent class regression and variable selection with applications to sleep pat- terns data. arXiv [Preprint]. 2025;2512.14903. https://doi.org/10.48550/arXiv.2512.14903.27. Messman BA, Wiley JF, Yap Y, Tung YC, Almeida IM, Dietch JR, et al. How much does sleep vary from night‐to‐night? A quantita- tive summary of intraindividual variability in sleep by age, gender, and racial/ethnic identity across eight‐pooled datasets. J Sleep Res. 2022;31:e13680. https://doi.org/10.1111/jsr.13680.28. Slavish DC, Taylor DJ, Lichstein KL. Intraindividual variability in sleep and comorbid medical and mental health conditions. Sleep. 2019;42:zsz052. https://doi.org/10.1093/sleep/zsz052.29. Veeramachaneni K, Slavish DC, Dietch JR, Kelly K, Taylor DJ, Elsevier. Intraindividual variability in sleep and perceived stress in young adults. Sleep Health. 2019;5:572–9. https://doi.org/10.1016/j.sleh.2019.07.009.30. Buysse DJ, Cheng Y, Germain A, Moul DE, Franzen PL, Fletcher M, et al. Night-to-night sleep variability in older adults with and without chronic insomnia. Sleep Med. 2010;11:56–64. https://doi.org/10.1016/j.sleep.2009.02.010.31. Chan WS, Williams J, Dautovich ND, McNamara JP, Stripling A, Dzierzewski JM, et al. Night-to-night sleep variability in older adults with chronic insomnia: mediators and moderators in a randomized controlled trial of brief behavioral therapy (BBT- I). J Clin Sleep Med. 2017;13:1243–54. https://doi.org/10.5664/jcsm.6790.32. Bei B, Wiley JF, Trinder J, Manber R, Elsevier. Beyond the mean: a systematic review on the correlates of daily intrain- dividual variability of sleep/wake patterns. Sleep Med Rev. 2016;28:108–24. https://doi.org/10.1016/j.smrv.2015.06.003.33. Annesley SJ, Missailidis D, Heng B, Josev EK, Armstrong CW, Elsevier. Unravelling shared mechanisms: insights from recent ME/CFS research to illuminate long COVID pathologies. Trends Mol Med. 2024;30:443–58. https:// doi. org/ 10. 1016/j. molmed.2024.02.001.34. Stussman B, Williams A, Snow J, Gavin A, Scott R, Nath A, et al. Characterization of post–exertional malaise in patients with ### Page 13 Journal of Clinical Sleep Medicine (2026) 22:77 Page 13 of 13 77myalgic encephalomyelitis/chronic fatigue syndrome. Front Neu- rol. 2020;11:1025. https://doi.org/10.3389/fneur.2020.01025.35. Hartle M, Bateman L, Vernon SD. Dissecting the nature of post- exertional malaise. Fatigue Biomed Health Behav. 2021;9:33–44. https://doi.org/10.1080/21641846.2021.1905415.36. Vernon SD, Hartle M, Sullivan K, Bell J, Abbaszadeh S, Unut- maz D, et al. Post-exertional malaise among people with long COVID compared to myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). Work. 2023;74:1179–86. https://doi.org/10.3233/WOR-220581.37. Irwin MR. Sleep and inflammation: partners in sickness and in health. Nat Rev Immunol. 2019;19:702–15. https://doi.org/10.1038/s41577-019-0190-z.38. Bonsall DR, Harrington ME. Circadian rhythm disruption in chronic fatigue syndrome. Adv Neuroimmune Biol. 2013;4:265– 74. https://doi.org/10.3233/NIB-130074.39. Smith AK, Fang H, Whistler T, Unger ER, Rajeevan MS. Con- vergent genomic studies identify association of GRIK2 and NPAS2 with chronic fatigue syndrome. Neuropsychobiology. 2011;64:183–94. https://doi.org/10.1159/000326692.40. McCarthy MJ. Circadian rhythm disruption in Myalgic Enceph- alomyelitis/Chronic Fatigue Syndrome: implications for the post-acute sequelae of COVID-19. Brain Behav Immun-Health. 2022;20:100412. https://doi.org/10.1016/j.bbih.2022.100412.41. Huang T, Kianersi S, Wang H, Potts KS, Noordam R, Sofer T, et al. Sleep duration irregularity and risk for incident cardiovascular disease in the UK Biobank. J Am Heart Assoc. 2025;14:e040027. https://doi.org/10.1161/JAHA.124.040027.42. Xu Y, Barnes VA, Harris RA, Altvater M, Williams C, Norland K, et al. Sleep variability, sleep irregularity, and nighttime blood pressure dipping. Hypertension. 2023;80:2621–6. https://doi.org/10.1161/HYPERTENSIONAHA.123.21497.43. Denu MKI, Revoori R, Eghan C, Kwapong FL, Hillman A, Nor- meshie CA, et al. Association between chronic fatigue syndrome/ myalgic encephalomyelitis and cardiovascular disease. Sci Rep. 2025;15:2294. https://doi.org/10.1038/s41598-025-86609-4.44. Natelson BH, Brunjes DL, Mancini D. Chronic fatigue syndrome and cardiovascular disease: JACC state-of-the-art review. J Am Coll Cardiol. 2021;78:1056–67. https://doi.org/10.1016/j.jacc.2021.06.045. American College of Cardiology Foundation Washington DC.45. Zerón-Rugerio MF, Zaragozá MC, Domingo JC, Sanmartín- Sentañes R, Alegre-Martin J, Castro-Marrero J, et al. Sleep and circadian rhythm alterations in myalgic encephalomyelitis/chronic fatigue syndrome and post-COVID fatigue syndrome and its asso- ciation with cardiovascular risk factors: A prospective cohort study. Chronobiol Int. 2024;41:1104–15. https://doi.org/10.1080/07420528.2024.2380020.Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.