Abstract
:Sleep is an important factor for recovery and performance in endurance sports, yet its role in ultra-endurance events remains unclear due to extreme physical and cognitive demands and disrupted sleep patterns. This systematic review aimed to analyze the role of sleep in physical and cognitive performance in ultra-endurance athletes.Methods:This systematic review followed PRISMA guidelines. A comprehensive search was conducted in May 2025 across PubMed/Medline, Embase, SPORTDiscus, and Web of Science. Two researchers independently
its role in ultra-endurance events remains unclear due to extreme physical and cognitive demands and disrupted sleep patterns. This systematic review aimed to analyze the role of sleep in physical and cognitive performance in ultra-endurance athletes.Methods:This systematic review followed PRISMA guidelines. A comprehensive search was conducted in May 2025 across PubMed/Medline, Embase, SPORTDiscus, and Web of Science. Two researchers independently screened, selected, extracted, and assessed data quality using the JBI tools (PROSPERO ID: CRD420251042220).Results:Of 424 articles, 16 met inclusion criteria, totaling data from 1389 athletes. Regarding physical performance, better outcomes were associated with no or less sleep during competition (TST), extended sleep the night before, and increased time in light sleep. In contrast, longer wake time, lower sleep quality, greater sleepiness during competition, and higher sleep efficiency were linked to poorer performance. Cognitive performance was positively associated with pre-race sleep quality and mid-race naps. Conversely, greater accumulated sleep before testing was linked to worse cognitive outcomes.Conclusions:Sleep, particularly total sleep time (TST), plays an important role in ultra-endurance performance, although this relationship may be non-linear and influenced by race context and individual strategies. Pre-race and intra-race sleep strategies such as napping and extended sleep may benefit performance. Further rigorous and longitudinal studies are needed to clarify sleep’s impact on performance and recovery in ultra-endurance contexts. Keywords:sleep; sleep quality; athletic performance; cognitive performance; ultra-endurance J. Clin. Med.2026,15, 1398 https://doi.org/10.3390/jcm15041398
J. Clin. Med.2026,15, 1398 2 of 17 1. Introduction The practice of ultra-endurance sports has exhibited exponential growth in recent years, contributing to a significant increase in the number of events dedicated to these disciplines [1,2]. These sports are defined by prolonged physical demands, typically lasting more than six hours and involving distances greater than 42.195 km (the standard marathon distance) [3,4]. Among the most commonly practiced ultra-endurance disciplines are ultramarathon running, long-distance triathlon, swimming, cycling, cross-country skiing, and adventure racing [3,4]. These disciplines place extreme demands on the human body [1], requiring prolonged exertion under challenging physiological and environmental conditions [5]. Such demands may lead to adverse health effects for participants [6,7], as they require a high level of both physical and cognitive performance. In this context, there is growing scientific interest in understanding the factors that influence performance in ultra-endurance disciplines, such age, sex, energy balance, body composition, maximal aerobic velocity, and nutritional behavior [1,8,9]. Sleep, however, has received little attention in this field, despite its well-established role in overall health and well-being. Adequate sleep is fundamental for the restoration of both physical and cognitive functions [10–13], and insufficient sleep duration and/or poor sleep quality (SQ) have been consistently associated with negative outcomes across multiple physiological systems, including the immune, cardiovascular, and musculoskeletal systems [14,15]. Furthermore, sleep disturbances can impair cognitive functions such as mood, alertness, and reaction time [10,16,17], which may significantly compromise athletic performance [18,19]. In sports, the relevance of sleep extends beyond general health. Studies have shown that athletes frequently experience poor sleep quality and insufficient sleep duration [18–20] , factors that can contribute to declines in physical performance, including reductions in strength, speed, and power, as well as cognitive impairments, such as mem- ory loss, decreased attention, and impaired decision-making [10]. These consequences are particularly critical for ultra-endurance athletes, who are exposed to high physiological demands over extended periods. To date, no studies have specifically considered sleep as a predictor of performance in ultra-endurance sports. Therefore, the present study aimed to analyze sleep-related characteristics and their associations with physical and cognitive performance in ultra-
loss, decreased attention, and impaired decision-making [10]. These consequences are particularly critical for ultra-endurance athletes, who are exposed to high physiological demands over extended periods. To date, no studies have specifically considered sleep as a predictor of performance in ultra-endurance sports. Therefore, the present study aimed to analyze sleep-related characteristics and their associations with physical and cognitive performance in ultra- endurance athletes. 2. Materials and Methods 2.1. Protocol and Registration This systematic review was conducted in accordance with the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [21]. The PRISMA checklist is available in Supplementary Table S1, and the study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) under the registration number CRD420251042220. 2.2. Eligibility Criteria Study eligibility was defined according to the PECOS framework (Population, Expo- sure, Comparators, Outcomes, Study design), which guided the inclusion criteria (Table). The central research question was: “Does better SQ and quantity improve the physical or cognitive performance of ultra-endurance athletes?” https://doi.org/10.3390/jcm15041398
J. Clin. Med.2026,15, 1398 3 of 17 Table 1.PECOS criteria adopted in this systematic literature review. PECOS Inclusion Criteria Participants Ultra-endurance athletes of both sexes (humans) Type of ultra-endurance sport considered: running, cycling, swimming, triathlon, rowing, skiing, or adventure racing. Exposition Sleep efficiency, sleep duration, sleep habits, sleep strategies, sleep disorders, and sleep quality Comparative or control Lower and higher sleep quality/quantity; short and long sleep duration Outcome measurement Physical or Cognitive performance in ultra-endurance events. Objective physical performance outcomes include: Time to complete; Final ranking or placement; Maximal oxygen uptake (VO2max); Muscular power; Time to exhaustion; Power output, average speed, and distance covered; Countermovement Jump Performance Reduction in performance compared to baseline. Cognitive performance outcomes may include: Reaction time; Decision-making ability; Mental confusion; Executive functions (e.g., working memory and cognitive flexibility); Attention; Psychomotor performance. Studies included Clinical trials, observational studies (cross-sectional, cohort), or case–control studies; relate case Eligible participants included ultra-endurance athletes of any gender and performance level (elite or amateur), who have participated in at least one ultra-endurance event. Eligible events encompassed running, cycling, swimming, triathlon, rowing, or skiing, whether single-stage or multi-stage, and must exceed the standard marathon distance (42.195 km) or last at least six hours. This includes off-road, track, treadmill, or road events. The criteria for non-inclusion or exclusion were as follows: (1) endurance events with distances shorter than 42.195 km or durations of less than six hours; (2) outcomes limited exclusively to physiological or biochemical markers (e.g., cortisol, cytokines, heart rate), with no correlation to performance; (3) assessments focused solely on mood, stress, anxiety, or motivation, without any association with cognitive function or performance; (4) publications that were not full original studies, such as abstracts, letters, case series, review articles, or studies for which full-text access was unavailable and (5) studies conducted in animal models. 2.3. Database and Search Strategy A systematic literature search was conducted in May 2025 independently and in paral- lel by two authors (LQG and BOR), using four major databases: PubMed/Medline, Embase, SPORTDiscus, and Web of Science. No restrictions were applied regarding publication date or language during the identification and
full-text access was unavailable and (5) studies conducted in animal models. 2.3. Database and Search Strategy A systematic literature search was conducted in May 2025 independently and in paral- lel by two authors (LQG and BOR), using four major databases: PubMed/Medline, Embase, SPORTDiscus, and Web of Science. No restrictions were applied regarding publication date or language during the identification and selection process. Search descriptors were selected based on terms from Medical Subject Headings (MeSH–PubMed), Health Sciences Descriptors (DeCS), and Emtree (Embase), all in English. https://doi.org/10.3390/jcm15041398
J. Clin. Med.2026,15, 1398 4 of 17 Three sets of indexed terms were used (1—population, 2—exposure, and 3—outcome), searched across all fields and combined using the Boolean operators “OR” and “AND” to appropriately link the search terms. Set 1 included terms related to ultra-endurance sports: (“ultra-endurance running” OR “ultra marathon” OR “ultramarathon running” OR “ultra-endurance” OR “ultra-athlete” OR “ultra-endurance training” OR “ultra-distance” OR “ultramarathon” OR “ultra-event” OR “trail run” OR “mountain run” OR “ultra run” OR “ultra trail” OR “ultra endurance” OR “cross country skiing” OR “cross country ski” OR “cross country skiers” OR “skating” OR “ironman” OR “triathlon” OR “bicycling” OR “ultraendurance sports” OR “treadmill exercise” OR “high intensity exercise”). Set 2 included different terminologies related to sleep: (“sleep” OR “sleeping habits” OR “sleep habits” OR “sleep duration” OR “sleep hygiene” OR “sleep quality” OR “sleep latency” OR “sleep stage” OR “sleep stages” OR “sleepiness” OR “sleep problem” OR “sleep disorder” OR “sleep–wake disorders” OR “sleep deprivation” OR “insufficient sleep” OR “sleep fragmentation” OR “fragmented sleep” OR “sleep loss” OR “restricted sleep” OR “sleep restriction” OR “sleep time” OR “night sleep” OR “sleep pattern” OR “sleep patterns” OR “somnolence” OR “sleep debt” OR “sleep spindle” OR “sleep spindles”) And set 3 included terms related to performance as the primary outcome: (“Physical Functional Performance” OR “Athletic Performance” OR “Physical performance” OR “Psychomotor Performance” OR “Cognitive performance” OR “Time to complete the race” OR “Final ranking” OR “Placement” OR “Maximal oxygen uptake” OR “VO2max” OR “Muscular power” OR “Time to exhaustion” OR “Power output” OR “Average speed” OR “Distance covered” OR “Reaction Time” OR “Response Speed” OR “Response Time” OR “Response Times” OR “Decision-Making” OR “Executive Function” OR “Executive Functions” OR “Cognitive Flexibility” OR “Cognitive Flexibilities” OR “Attention” OR “Focus of Attention” OR “Attention Focus” OR “Decision-making ability” OR “Mental confusion” OR “Working memory”). The complete search strategy for each database is presented in Supplementary Table S2. 2.4. Data Extraction and Selection Process Initially, all retrieved records were imported into Zotero software (version 7.0.15; Corporation for Digital Scholarship, Vienna, Austria), which assists in organizing references and identifying duplicate entries for subsequent
Attention” OR “Attention Focus” OR “Decision-making ability” OR “Mental confusion” OR “Working memory”). The complete search strategy for each database is presented in Supplementary Table S2. 2.4. Data Extraction and Selection Process Initially, all retrieved records were imported into Zotero software (version 7.0.15; Corporation for Digital Scholarship, Vienna, Austria), which assists in organizing references and identifying duplicate entries for subsequent removal [22]. Following the removal of duplicates, the remaining records were imported into the Rayyan QCRY ® platform (Qatar Computing Research Institute, Doha, Qatar) [23], where we screened the papers based on title and abstract review, followed by full-text reading to assess the eligibility of the studies. All procedures were carried out independently, in parallel, and in a blinded manner by two reviewers (LQG and BOR) to ensure greater reliability in the process. In cases of disagreement during the title and abstract screening phase, the study was automatically included for full-text review. When discrepancies arose after full-text reading, they were resolved by consensus between the two reviewers (LQG and BOR), or, if consensus was not reached, by consulting a third reviewer (HdSS). After the full screening of studies, data extraction was performed using a summary table created by the authors (LQG and BOR). This table included the following relevant in- formation: (1) study reference (author and year of publication); (2) study design; (3) sample characteristics (number of participants, sex, age); (4) exposure characteristics (sleep-related variables and sleep assessment methods); (5) characteristics of the ultra-endurance sport (ultramarathon, triathlon, cycling, swimming, rowing, skiing); (6) performance variables https://doi.org/10.3390/jcm15041398
J. Clin. Med.2026,15, 1398 5 of 17 assessed (physical or cognitive, along with the methods used to assess performance); and (7) main findings. Data extraction was conducted using a standardized form in Microsoft ® Excel ® 2019 (version 16.0; Microsoft Corp., Redmond, WA, USA), which contained the variables of interest (Table). Data were independently extracted by the authors (LQG and BOR) and subsequently verified for consistency. Any discrepancies were resolved by a third author (HdSS). 2.5. Study Quality Assessment The risk of bias assessment was performed independently and in parallel by two researchers (LQG and BOR), with any discrepancies resolved by consensus. This critical appraisal was conducted using the Joanna Briggs Institute (JBI) Reviewer’s Manual (Joanna Briggs Institute, Adelaide, SA, Australia) and its specific Critical Appraisal Tools tailored to each study design [24] (Figure). Three JBI checklists were applied according to the design of the studies evaluated: the Checklist for Cohort Studies, which includes 11 items; the Checklist for Analytical Cross- Sectional Studies, containing 8 items; and the Checklist for Case Reports, also comprising 8 items. Each item was assessed as “met = yes,” “not met = no,” or “unclear,” and in some cases, as “not applicable.” The JBI manual recommends that authors predefined criteria to classify the risk of bias level for each study, since the tool does not provide a standardized scoring system to determine whether the risk of bias is low, moderate, or high for each article (JBI Manual for Evidence Synthesis, 2024) [25]. To classify each article individually, therefore, the following cut-off points were adopted according to the percentage of affirmative responses: low (≥70%), moderate (between 50 and 70%) and high risk of bias (<50%) [26]. (A1) Figure 1.Cont. https://doi.org/10.3390/jcm15041398
J. Clin. Med.2026,15, 1398 6 of 17 (A2) (A3) Figure 1.Risk of bias assessment according to the Joanna Briggs Institute: (A1) Cohort Studies, (A2) Analytical Cross-Sectional Studies, and (A3) Case Reports [27–42]. 2.6. Risk of Bias The assessment of risk of bias is summarized in Figure, according to the type of study evaluated. FigureA1 presents the analysis of the eight prospective studies: two were classified as having low risk of bias, five as moderate risk, and one as high risk. The https://doi.org/10.3390/jcm15041398
J. Clin. Med.2026,15, 1398 7 of 17 studies classified as moderate risk frequently relied on subjective or non-validated methods to assess sleep exposure and did not adequately address potential confounding factors. The only study rated as high risk [27] failed to provide sufficient follow-up and did not report strategies to deal with incomplete data, substantially increasing its overall risk. FigureA2 displays an assessment of the five cross-sectional studies: two were judged to have low risk of bias, and three moderate risk. The main reasons for moderate risk in this group included the use of non-validated sleep measures, lack of standardized outcome criteria, and insufficient reporting of strategies to control for confounding factors. For instance, Sinisgalli et al. [28] and Poussel et al. [29] both showed high risk in the validity of outcome measurement, while Martin et al. [30] did not address confounders in the analysis. FigureA3 shows the analysis of the three case reports, all of which were classified as having moderate risk of bias. In these studies, the domain most commonly associated with increased bias was the lack of identification or reporting of adverse events (harms) or unexpected outcomes. Specifically, Anderson et al. [31] also presented unclear reporting of patient history, while Bianchi et al. [32] and Biorci et al. [33] did not describe harms, contributing to their moderate risk classification. 3. Results 3.1. Study Selection and Characteristics Initially, a total of 424 studies were identified through searches in the databases (PubMed/Medline, Embase, SPORTDiscus, and Web of Science) in May 2025. Subsequently, 112 duplicate records were removed, leaving 312 articles for title and abstract screening. At this stage, 279 studies were excluded for not meeting the inclusion criteria. As a result, 34 articles were considered for full-text review, of which 16 met the eligibility criteria and were included in this systematic review (Figure). The studies excluded at this final stage are listed with their respective reasons for exclusion in Supplementary Table S3. Figure 2.PRISMA 2020 flow diagram illustrating the study selection process. https://doi.org/10.3390/jcm15041398
which 16 met the eligibility criteria and were included in this systematic review (Figure). The studies excluded at this final stage are listed with their respective reasons for exclusion in Supplementary Table S3. Figure 2.PRISMA 2020 flow diagram illustrating the study selection process. https://doi.org/10.3390/jcm15041398
Description
Analyzes sleep's impact on performance in ultra-endurance athletes.