Fatigue is measured across a wide spectrum of methods, from simple questionnaires that ask how tired you feel to laboratory biomarkers that track molecular changes in your blood, brain, and muscles. No single measurement captures the full picture, because fatigue itself is not one thing: it can originate in the brain, in the muscles, in the immune system, or in some combination of all three. The most reliable assessments tend to combine subjective self-reports with at least one objective measure, and the field is rapidly moving toward wearable sensors and machine-learning models that fuse multiple data streams in real time.
Self-Report Questionnaires
The simplest and most widely used fatigue measurements are standardized questionnaires. The Multidimensional Fatigue Inventory (MFI-20) is among the most established. It asks 20 questions across five proposed dimensions: general fatigue, physical fatigue, reduced activity, reduced motivation, and mental fatigue. The original validation work found good internal consistency and supported the five-factor structure across multiple samples.1PubMed. The Multidimensional Fatigue Inventory (MFI) psychometric qualities of an instrument to assess fatigue Other commonly used tools include the Fatigue Severity Scale, the Chalder Fatigue Scale, and the Karolinska Sleepiness Scale, each designed with a slightly different emphasis.
Questionnaires are cheap, fast, and require no equipment. But they have real limitations. A large German population study that re-examined the MFI-20’s five-factor structure found it did not hold up well statistically, with fit indices falling below acceptable thresholds. Instead, the data suggested that most of the variation could be explained by just one or two underlying factors rather than five distinct dimensions.2PubMed Central. Psychometric properties, factor structure, and German population norms of the multidimensional fatigue inventory (MFI-20) That matters because if a questionnaire claims to separate mental fatigue from physical fatigue but the scores move together regardless, you may be measuring the same underlying state twice. Self-reports are also vulnerable to mood, context, and individual interpretation of vague words like “tired” or “drained.” They remain a necessary starting point, but researchers increasingly treat them as one input among several rather than the final word.
Reaction Time Testing
When fatigue affects alertness, it slows your ability to respond to simple stimuli, and that slowing can be measured precisely. The Psychomotor Vigilance Test (PVT) is the gold standard for this. You watch a screen and press a button as soon as a light appears. As fatigue accumulates from sleep loss or extended wakefulness, reaction times get slower and “lapses” (responses delayed beyond a threshold) become more frequent. The PVT reliably captures fatigue-related changes in alertness from both total and partial sleep deprivation, with large effect sizes for response speed under total sleep deprivation.3PubMed Central. Validity and Sensitivity of a Brief Psychomotor Vigilance Test (PVT-B) to Total and Partial Sleep Deprivation
A shorter version of the test (the PVT-B, lasting about three minutes instead of ten) shows comparable sensitivity when the analysis accounts for how timing between stimuli affects the results. Research has found that using only longer intervals between stimuli produces stronger effect sizes for fatigue, because short intervals introduce noise unrelated to tiredness.4PubMed. Using interstimulus interval to maximise sensitivity of the Psychomotor Vigilance Test to fatigue The PVT is widely used in sleep research, military operations, and transportation safety, and its main limitation is that it captures fatigue-related sleepiness specifically. If your fatigue is more about muscle weakness or motivational depletion than about drowsiness, the PVT may not pick it up.
Heart Rate Variability
Your heart does not beat at a perfectly steady rhythm. The tiny fluctuations between beats, called heart rate variability (HRV), reflect how well your autonomic nervous system is balancing its “fight or flight” and “rest and digest” branches. When you are fatigued, that balance shifts, and HRV tends to drop. A study of healthy young adults found that several HRV measures declined significantly after a fatiguing physical task, indicating reduced autonomic adaptability.5PubMed Central. Effects of Induced Physical Fatigue on Heart Rate Variability in Healthy Young Adults
HRV has also been tested as a fatigue marker in driving contexts, where staying alert is a safety concern. Structural equation modeling found that the HRV measure known as RMSSD (which reflects short-term beat-to-beat variability) negatively predicted fatigue, while the ratio of low-frequency to high-frequency HRV power positively predicted it. Lower RMSSD and a higher ratio were both associated with greater fatigue, though HRV explained a relatively modest share of the overall variance in fatigue levels.6Transportation Research Part F: Traffic Psychology and Behaviour. Heart rate variability as an indicator of fatigue: A structural equation model approach In people with chronic fatigue syndrome, the connection appears stronger: patients showed decreased HRV across multiple measures compared to healthy controls, and self-reported fatigue severity correlated with specific HRV indices.7PubMed Central. Reduced heart rate variability predicts fatigue severity in individuals with chronic fatigue syndrome/myalgic encephalomyelitis
HRV is appealing because it can be tracked continuously with a chest strap or even a smartwatch. But it is influenced by caffeine, hydration, temperature, fitness level, and emotional state, so interpreting a single reading in isolation is unreliable. Tracking trends over days or weeks within the same individual tends to be more informative than comparing one person’s numbers to another’s.
Brain Activity Measured by EEG
Electroencephalography (EEG) picks up electrical activity from the scalp and can reveal characteristic shifts in brain-wave patterns as mental fatigue sets in. A systematic review with meta-analyses found large increases in theta-band activity across frontal, central, and posterior brain regions during mental fatigue, making theta power a robust biomarker. Alpha-band activity also increased, though the effect was moderate and more variable between individuals.8PubMed. The influence of mental fatigue on brain activity: Evidence from a systematic review with meta-analyses In plain terms, theta waves are associated with drowsy or unfocused states, and their consistent rise during prolonged cognitive tasks makes them one of the more dependable objective markers researchers have.
A related technique, functional near-infrared spectroscopy (fNIRS), measures blood oxygenation changes in the brain’s surface layers. A meta-analysis of fNIRS studies concluded that mental fatigue causes significant activation of the prefrontal cortex and that fNIRS is an effective tool for tracking that activation in real time.9PubMed. Mental fatigue causes significant activation of the prefrontal cortex: A systematic review and meta-analysis of fNIRS studies Interestingly, in people with traumatic brain injury who report chronic mental fatigue, the pattern reverses: the fatigued group showed lower oxygenated hemoglobin in prefrontal regions from the start of testing and a reduced ability to ramp up prefrontal activity when the task became harder.10PubMed Central. Mental Fatigue and Functional Near-Infrared Spectroscopy (fNIRS) – Based Assessment of Cognitive Performance After Mild Traumatic Brain Injury This distinction matters clinically: in a healthy brain, fatigue shows up as the prefrontal cortex working harder to compensate; in an injured brain, the compensatory mechanism itself may be impaired.
Sleep Architecture
Fatigue and poor sleep are so intertwined that sleep studies often serve as indirect fatigue measures. Polysomnography, which records brain waves, eye movements, and muscle tone during sleep, can reveal disruptions that explain daytime fatigue even when total sleep duration looks normal. In chronic fatigue syndrome, sleep studies have found increased fragmentation and differences in slow-wave sleep. Spectral analysis showed that reduced power in the ultra-slow frequency range during deep sleep correlated with both fatigue severity and impaired sleep quality.11PubMed. Slow wave sleep in the chronically fatigued: Power spectra distribution patterns in chronic fatigue syndrome and primary insomnia In other words, even when people with chronic fatigue appear to be in deep sleep, the quality of that sleep may be subtly degraded in ways that standard staging criteria miss.
Muscle Fatigue and Surface EMG
When fatigue originates in the muscles rather than the brain, surface electromyography (EMG) can quantify it. Sensors placed on the skin over a muscle pick up its electrical activity during contractions. As a muscle fatigues, the median frequency of that electrical signal drops, because fatigued muscle fibers fire more slowly. Researchers tracking shoulder muscles during repeated arm raises found significant load-related drops in median-frequency slopes, with the upper trapezius showing greater fatigue responses than the lower trapezius.12PubMed Central. Continuous wavelet transform-based analysis of shoulder muscle fatigue during dynamic abduction: a surface electromyography study Interestingly, the non-dominant side showed larger frequency drops in some muscles, suggesting that hand dominance affects how quickly peripheral fatigue develops.
A long-running debate in exercise science is whether what looks like “central” fatigue (originating in the brain’s drive to the muscles) is actually peripheral fatigue in disguise. Simulation studies have shown that force declines commonly attributed to the central nervous system can be explained entirely by peripheral factors, and that the traditional method of testing for central fatigue using electrical stimulation has important flaws.13PubMed Central. Is the notion of central fatigue based on a solid foundation? This does not mean central fatigue does not exist, but it does mean that cleanly separating where fatigue “lives” in the body is harder than textbooks suggest.
Inflammatory and Immune Markers
In conditions like chronic fatigue syndrome (CFS, also called ME/CFS), fatigue is not just a feeling. It correlates with measurable changes in the immune system. A study of ME/CFS patients found that 17 cytokines showed a statistically significant upward trend that correlated with disease severity, and 13 of those 17 were pro-inflammatory.14PubMed Central. Cytokine signature associated with disease severity in chronic fatigue syndrome patients Separate work confirmed that higher levels of specific pro-inflammatory cytokines, including GM-CSF and TNF-alpha, correlated with greater physical symptoms and fatigue in ME/CFS patients.15PubMed Central. Proteomics and cytokine analyses distinguish myalgic encephalomyelitis/chronic fatigue syndrome cases from controls Another study linked increased IL-1 and TNF-alpha specifically to fatigue, autonomic symptoms, and flu-like malaise in ME/CFS.16PubMed. Evidence for inflammation and activation of cell-mediated immunity in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS): increased interleukin-1, tumor necrosis factor-α, PMN-elastase, lysozyme and neopterin
These findings are significant because they suggest fatigue in ME/CFS has a biological immune signature, not just a psychological one. But cytokine profiles are expensive to measure, vary with time of day and recent activity, and differ between studies. They are research tools at this stage, not something your doctor would routinely order to evaluate tiredness.
Cortisol and the Stress Axis
Cortisol, the body’s main stress hormone, follows a natural daily rhythm: it peaks shortly after waking and declines through the evening. In people with chronic fatigue syndrome, that rhythm is often flattened. Studies using saliva samples collected across the day have found lower cortisol in the morning and higher cortisol in the evening compared to healthy controls, resulting in a blunted daily curve.17Psychosomatic Medicine. Alterations in Diurnal Salivary Cortisol Rhythm in a Population-Based Sample of Cases With Chronic Fatigue Syndrome Separate work measuring salivary cortisol and cortisone confirmed that levels were lower across the whole day in CFS patients, though the timing of the daily peak was not shifted.18PubMed. Diurnal patterns of salivary cortisol and cortisone output in chronic fatigue syndrome Urinary cortisol measurements showed the same pattern of reduced levels throughout the daily cycle.19PubMed. Diurnal excretion of urinary cortisol, cortisone, and cortisol metabolites in chronic fatigue syndrome
A flattened cortisol curve is not unique to CFS. It shows up in burnout, depression, and chronic stress as well, so it is more of a general marker of disrupted stress regulation than a fatigue-specific diagnostic. Still, tracking cortisol over a full day using saliva samples is one of the more accessible hormonal assessments and gives clinicians a window into how the body’s stress system is functioning.
Eye Tracking and Pupil Responses
Your eyes betray fatigue in several measurable ways. Blink rate, eyelid closure percentage, and pupil size all change as alertness declines. A systematic review found that tonic pupil size, controlled by competing signals from the sympathetic and parasympathetic nervous systems, has the potential to indicate mental fatigue under controlled conditions, though blink-based measures borrowed from sleep research should be used with caution in fatigue contexts.20PubMed. Mental fatigue measurement using eye metrics: A systematic literature review Researchers have also explored automated pupillometry, which measures the pupil’s light reflex, as a possible fatigue detection tool in driving scenarios. Combining pupil light reflex data with HRV data showed promise for detecting driving fatigue more accurately than either measure alone.21PubMed Central. Assessment of Combination of Automated Pupillometry and Heart Rate Variability to Detect Driving Fatigue
Wearable Sensors in the Workplace
Outside the lab, wearable devices are increasingly used to track fatigue in real work environments. A study of construction workers used wrist-worn heart rate sensors combined with a bioenergetic model to estimate whole-body fatigue, achieving a correlation of 0.83 with workers’ self-reported fatigue and a mean error of about 13%.22PubMed. Wearable heart rate sensing and critical power-based whole-body fatigue monitoring in the field Another approach used smartwatches capturing pulse rate, skin conductance, skin temperature, and motion data from industrial workers. When machine-learning models incorporated contextual information like shift timing and task load alongside the biometric data, classification accuracy improved substantially, reaching an F1 score above 0.93 for binary fatigue detection.23Safety Science. Early detection of physical fatigue in industry using wearable sensors and contextual modeling
The practical value here is obvious for safety-critical industries like construction, mining, and transportation, where undetected fatigue leads to accidents. But these systems raise questions about worker privacy and the reliability of algorithms trained on one workforce being applied to another. Personalization seems to matter: models that account for individual baseline physiology and job context consistently outperform generic ones.
Keystroke Dynamics and Passive Digital Sensing
One of the more unexpected frontiers in fatigue measurement involves the way you type. As mental fatigue accumulates over hours of wakefulness, subtle changes appear in typing speed and rhythm. A study tracking natural smartphone use found that typing speed follows an inverted U-shape across the day: it increases after waking, peaks at roughly seven to eight hours of wakefulness, then declines to below-average levels by about fifteen hours awake.24PLOS Digital Health. Patterns of smartphone typing performance by time awake: implications for unobtrusive ambulatory mental fatigue assessment The appeal of keystroke dynamics is that they require no special equipment and no conscious effort from the user; the phone or computer simply analyzes typing patterns in the background.
Feasibility research in the general population confirmed that psychomotor patterns characteristic of mental fatigue do manifest in natural typing and can be detected by automated analysis.25PubMed Central. Detection of Mental Fatigue in the General Population: Feasibility Study of Keystroke Dynamics as a Real-world Biomarker In people with multiple sclerosis, keystroke dynamics also showed potential as a remote cognitive monitoring tool, which could help detect changes in functioning without requiring clinic visits.26PubMed Central. Associations between smartphone keystroke dynamics and cognition in MS This is still early-stage work, and the challenge is separating fatigue-driven typing changes from other influences like distraction, multitasking, or simply using a different app. But as a low-burden complement to other measures, it is a genuinely novel direction.
Voice-Based Fatigue Detection
Speech changes when you are fatigued. Pitch, rhythm, articulation clarity, and energy all shift in ways that are difficult to fake. Researchers have built machine-learning models that extract acoustic features from voice samples and match them to fatigue levels. One system achieved about 94% accuracy when classifying fatigue from short voice samples and about 81% from longer stretches of speech.27PubMed Central. A rapid, non-invasive method for fatigue detection based on voice information Like keystroke analysis, voice-based assessment is attractive because it could be embedded into everyday technology, such as a phone app or a vehicle communication system, without requiring the user to stop and take a test.
Molecular Markers in the Blood
At the molecular level, researchers are exploring whether blood-based biomarkers can reliably distinguish fatigued from non-fatigued individuals. One active area involves microRNAs (miRNAs), small molecules that help regulate gene expression. In ME/CFS, several studies have identified sets of circulating miRNAs that differ between patients and healthy controls. One study found six miRNAs that were differently expressed in patients’ plasma, with five upregulated and one downregulated, and their levels correlated with disease severity.28PubMed Central. Circulating miRNAs Expression in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome Another identified 15 miRNAs with acceptable discriminatory capacity between ME/CFS and healthy controls based on area-under-the-curve values above 0.75.29Scientific Reports. Assessing diagnostic value of microRNAs from peripheral blood mononuclear cells and extracellular vesicles in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome A machine-learning approach using 11 circulating miRNAs achieved perfect classification between ME/CFS and fibromyalgia in one dataset, though such results need replication in larger, independent cohorts before they can be considered diagnostic tools.30Scientific Reports. Circulating microRNA expression signatures accurately discriminate myalgic encephalomyelitis from fibromyalgia and comorbid conditions
Metabolomics, which profiles hundreds of small molecules in the blood at once, offers another window. In women with chronic widespread pain, lower levels of the omega-3 fatty acid EPA were associated with greater fatigue, and other metabolites in lipid and amino acid pathways also showed significant links.31PubMed Central. Metabolomic markers of fatigue: Association between circulating metabolome and fatigue in women with chronic widespread pain Animal studies have pointed to changes in branched-chain amino acids and urea-cycle metabolites as markers of physical fatigue.32PLoS ONE. Potential Biomarkers of Fatigue Identified by Plasma Metabolome Analysis in Rats These metabolomic approaches are promising but require specialized laboratory analysis and are far from clinical use.
The Gut Microbiome Connection
An emerging line of research links fatigue severity to the composition of gut bacteria. In CFS patients, certain bacterial families associated with producing short-chain fatty acids were depleted, and the abundance of those bacteria correlated with fatigue scores: lower levels of Rikenellaceae and Ruminococcaceae were tied to worse fatigue.33Scientific Reports. Alterations in gut microbiota and associated metabolites in patients with chronic fatigue syndrome Broader reviews have proposed that microbiome disruption and the resulting “metabolic endotoxemia” may themselves function as disorder biomarkers, with the loss of beneficial gut organisms setting off a cascade that affects plasma metabolite levels and ultimately drives fatigue symptoms.34Microbes & Immunity. Microbial involvement in myalgic encephalomyelitis/chronic fatigue syndrome pathophysiology This field is still young and the findings are correlational. Nobody yet has a gut-bacteria test that can diagnose or quantify fatigue. But it reinforces the broader point that fatigue is a whole-body phenomenon, not just a brain state or a muscle state.
Combining Multiple Measures
Given how many systems fatigue touches, it is not surprising that combining data streams outperforms any single measure. In air traffic control research, a deep-learning model that fused EEG signals with eye movement data recognized mental fatigue more accurately than either input alone.35Advanced Engineering Informatics. Air traffic controllers’ mental fatigue recognition: A multi-sensor information fusion-based deep learning approach The industrial wearable study described earlier saw a similar jump when contextual information was layered on top of physiological sensors.23Safety Science. Early detection of physical fatigue in industry using wearable sensors and contextual modeling This multimodal trend reflects a growing consensus that fatigue is not well served by a single number or a single test. A questionnaire captures the subjective experience; HRV captures autonomic tone; EEG captures brain-state shifts; and reaction-time tests capture behavioral consequences. Each gives a partial view, and the question researchers are working on now is how to integrate them efficiently enough for real-world use without burying a person in sensors.
For someone trying to track their own fatigue, the practical takeaway is that a daily self-report paired with one objective measure, whether that is HRV from a smartwatch, reaction-time scores from a phone app, or even habitual typing speed, will give a far richer picture than either alone. For clinicians assessing patients with chronic fatigue conditions, the inflammatory, hormonal, and molecular markers discussed above are converging toward a future where a blood panel could help confirm and quantify what patients have long described. That future is not here yet, but the distance between “I just feel exhausted” and “here is the biology behind it” is closing faster than at any point in the history of fatigue research.