INTRODUCTION
Circadian rhythms regulate sleep and wakefulness, but their clinical relevance is not limited to sleep. They are linked to metabolism, cognition, mood, immune function, and tissue physiology. Circadian disruption can be a consequence of disease, a contributing factor in disease development, and a modifier of disease severity or progression [1]. Three basic concepts are useful for interpreting circadian measures. Phase refers to the timing of a rhythm, such as the timing of melatonin onset, sleep onset, or peak activity. Period refers to the length of one complete circadian cycle, which is normally synchronized to the 24-hour day but may differ across individuals under controlled conditions. Amplitude refers to the strength or magnitude of rhythmic oscillation, and reduced amplitude may indicate a weakened or fragmented rhythm.
In clinical practice, circadian rhythm assessment remains limited. Chronotype questionnaires, sleep diaries, actigraphy, and dim light melatonin onset (DLMO) are commonly used tools, but they do not capture the same aspect of circadian biology. Chronotype reflects a behavioral tendency, sleep diaries describe reported sleep-wake timing, actigraphy captures rest-activity patterns in daily life, and DLMO estimates internal circadian phase under controlled light conditions [2,3]. However, similar sleep-wake patterns can reflect different temporal mechanisms. Delayed sleep timing, for example, may result from a delayed internal phase, evening light exposure, social schedule, depressive symptoms, reduced daytime activity, or medication effects. These mechanisms are not distinguishable by sleep timing alone. Thus, circadian assessment should not rely on a single type of measurement.
Recent methods have expanded the type of circadian information that can be obtained from patients. Wearable sensors can measure daily rest-activity patterns, light exposure, physical activity, and heart-rate rhythms. Patient-derived cells can be used to study cellular circadian period under controlled ex vivo conditions. Blood transcriptomics, metabolomics, and proteomics may estimate internal biological time or characterize time-dependent molecular state [4-6].
This review proposes a practical framework for interpreting circadian information obtained from patients. Trait-like and state-like properties are not strictly dichotomous, but they remain useful for interpretation. Wearable-derived rest-activity rhythms mainly reflect the patient’s current rhythm state in daily life. DLMO and other controlled in vivo phase markers estimate internal circadian phase. Patient-derived cellular period measured under controlled ex vivo conditions is closer to an endogenous, trait-like circadian property. Blood-based omics may estimate molecular body time, but it also reflects systemic biological state. This practical classification may help avoid two errors: assuming that all biological samples represent stable traits and interpreting wearable-or sensor-derived rhythms as direct measures of internal circadian phase or endogenous circadian properties.
CIRCADIAN MEASURES OBTAINABLE FROM PATIENTS
The concepts of phase, period, and amplitude are useful, but they cannot be interpreted in isolation from the method used to measure them. Therefore, the practical question is not only whether a circadian rhythm is present, but also what aspect of circadian biology a given measure primarily represents. Circadian information obtained from patients can be grouped according to what each measure primarily reflects. This grouping should not be based only on the source of measurement. Wearable data, physiological assays, patient-derived cell models, and blood-based omics profiles may all provide circadian information, but they answer different questions. Some measures describe the patient’s current rhythm state in daily life. Some estimate internal circadian phase. Some measure cellular circadian period under controlled conditions. Others estimate molecular body time while also reflecting systemic biological state. These types of circadian information and their interpretation are summarized in Table 1.
The first type of information is environmental zeitgeber exposure. Light exposure, physical activity timing, meal timing, social schedule, and sleep environment provide external time cues to the circadian system [7]. These measures are usually state-like and modifiable. They do not measure the endogenous clock, but they help explain why a patient’s rhythm is advanced, delayed, weak, or irregular. Light exposure is especially important because it can reset human circadian phase, and its effect depends on timing, duration, intensity, and spectral properties [8,9].
The second type is behavioral rhythm state. Sleep-wake timing and actigraphy-derived rest-activity patterns describe how the patient’s rhythm is expressed in daily life. Common measures include acrophase, rhythm amplitude, interdaily stability, intradaily variability, M10, L5, and relative amplitude. These variables are clinically useful because they reflect real-world rhythm expression. However, they are influenced by behavior, environment, disease, medication, and caregiving or institutional routines. Large actigraphy studies show that rest-activity rhythm metrics vary by age, demographic factors, lifestyle, and health status, supporting their interpretation as state-sensitive rhythm markers rather than direct measures of the endogenous clock [10,11]. In patients with cognitive impairment, actigraphy-derived rest-activity patterns have also been associated with amyloid burden, medial temporal lobe atrophy, and cognitive function, suggesting that real-world rhythm state may be clinically informative in neurodegenerative conditions [12].
The third type is internal circadian phase. DLMO remains the most established marker of human circadian phase, but core body temperature rhythm and cortisol rhythm can also provide phase-related information. These measures are closer to internal circadian timing than sleep timing or actigraphy. However, they require careful sampling conditions and should be interpreted in relation to recent light exposure, sleep timing, and behavior. Experimental human studies show that changes in sleep-wake schedule and light exposure can shift phase and alter amplitude across multiple circadian outputs, including melatonin, cortisol, and body temperature rhythms [13].
The fourth type is endogenous cellular circadian property. For example, patient-derived fibroblasts, induced pluripotent stem cell (iPSC)-derived cells, or directly converted neurons can be used to measure cellular circadian period and rhythm characteristics under controlled ex vivo conditions. This information is closer to a trait-like property than wearable-derived rhythm state. However, trait-like does not mean purely genetic or fixed. Human fibroblast period varies across individuals, and physiological period measured in vivo has been reported to relate to fibroblast period in the same subjects [5,14]. As one example, our group recently measured cellular circadian period and its deviation from 24-hour cycle in cognitively impaired older adults. These measures were associated with Alzheimer-related pathology, brain aging markers, cognitive function, and clinical progression, suggesting that cellular period and its gap from the 24-hour cycle may provide endogenous circadian information relevant to neurodegenerative disease research [15]. A recent commentary article also discussed the broader significance of this approach for neurodegenerative disease biology [16].
The fifth type is molecular body time. Blood transcriptomics, metabolomics, proteomics, and related molecular approaches may estimate internal biological time. This information is different from cellular period. Blood-based omics may reflect circadian phase, but it also reflects systemic biological state, including immune activation, sleep loss, feeding, medication, and disease-related changes. Several approaches have been proposed to estimate internal body time from blood transcriptomic or metabolomic data [17-19]. However, blood molecular rhythms are also sensitive to sleep restriction and sleep deprivation, supporting the view that blood-based omics should be interpreted as both phase-related and state-sensitive [20,21].
These types of information can be compared to characterize circadian misalignment across different measures. Examples include the difference between cellular period and 24-hour cycle, the mismatch between internal phase and sleep-wake schedule, and the mismatch between trait-like circadian tendency and current rhythm state. These constructs are not direct biomarkers. They are interpretive tools that may help connect circadian assessment to individualized intervention (Figure 1).
FROM CIRCADIAN MEASURES TO CIRCADIAN MISALIGNMENT
The value of using multiple circadian measures is not simply to collect more variables. Its main value is to compare measures that reflect different aspects of circadian biology. A single measure may show delayed sleep timing, fragmented rest-activity rhythm, altered internal phase, or abnormal molecular timing. However, comparison across measures may help clarify whether the main problem is related to environmental timing, internal phase, endogenous period, systemic biological state, or discordance between biological timing and daily schedule. The concept of circadian misalignment is useful here because it includes several forms of discordance, including misalignment between sleep-wake timing and biological night, between feeding rhythm and sleep-wake rhythm, and between central and peripheral rhythms [1,22].
One example is period-related misalignment. Human circadian rhythms are normally entrained to the 24-hour day, but the intrinsic circadian period is not exactly 24 hours in all individuals. Patient-derived cellular period may provide one way to estimate an endogenous period-related property under controlled conditions. The difference between cellular period and 24-hour cycle can therefore be considered a potential indicator of period-related strain or vulnerability. This interpretation should remain cautious because cellular period is not a direct measure of the central circadian pacemaker. However, it provides a testable hypothesis: patients with a larger deviation from 24-hour cycle may require stronger, more regular, or more precisely timed zeitgeber input to maintain stable alignment with the external day.
A second example is phase-related misalignment. Internal circadian phase, estimated by DLMO or other controlled in vivo phase markers, may not match sleep-wake timing. A patient may sleep late because the internal phase is delayed. Another patient may sleep late because of evening light exposure, social schedule, depressive symptoms, low daytime activity, or medication effects. Conversely, a patient may keep an apparently regular sleep-wake schedule while the internal phase remains misaligned. Thus, phase-related misalignment cannot be inferred from sleep timing alone. For example, a larger mismatch between DLMO and habitual sleep timing may be tested as a predictor of sleep disturbance, daytime dysfunction, or response to phase-based intervention.
A third example is zeitgeber-related misalignment. Light exposure, physical activity, meals, and social contact are not only behaviors or environmental features. They also act as timing cues. Low daytime light exposure, excessive evening light, irregular meals, and fragmented daytime activity can weaken or shift circadian alignment. Feeding time is particularly relevant to peripheral clocks because peripheral tissues can be strongly affected by metabolic and feeding-related cues, sometimes in a way that diverges from the central clock. This creates the possibility of internal misalignment between central and peripheral circadian systems [22,23]. For example, irregular light exposure, activity timing, or meal timing may be tested as predictors of fragmented rest-activity rhythms and as modifiable targets for rhythm stabilization.
A fourth example is central-peripheral or tissue-level misalignment. The circadian system is not a single clock. It includes the central pacemaker in the suprachiasmatic nucleus and peripheral clocks in multiple organs, including metabolic tissues and the cardiovascular system. These clocks normally maintain coordinated phase relationships, but their alignment can be disturbed when light, sleep-wake timing, feeding, activity, or medication schedules provide conflicting timing signals. In cardiovascular physiology, for example, the heart has local circadian regulation, while also receiving neural, hormonal, and behavioral timing inputs from the central system. Misalignment between these systems may be relevant to cardiovascular function and chronotherapy [24,25]. For example, discordance between sleep-wake timing and feeding or medication timing may be tested in relation to metabolic, cardiovascular, or inflammatory outcomes.
A fifth example is trait-state misalignment. A patient may have a trait-like biological tendency, such as a longer cellular period, but live under a schedule that requires early activity. Another patient may have a relatively stable cellular period but show fragmented wearable-derived rhythm because of inactivity, illness, depressive symptoms, or care environment. In both cases, the discordance between biological tendency and current rhythm state may be more informative than either measure alone. For example, mismatch between a trait-like circadian tendency and the required daily schedule may be tested as a predictor of rhythm instability, impaired daytime performance, and depressive symptoms in workers.
Blood-based molecular body time may add another comparison. Molecular body time can be compared with clock time, sleep-wake timing, DLMO, or rest-activity rhythm. However, blood omics is also state-sensitive. A molecular estimate that appears phase-delayed may reflect true circadian phase, but it may also reflect sleep loss, inflammation, altered feeding, medication, or changes in blood cell composition. Therefore, molecular body time should be interpreted with behavioral and clinical context. For example, discrepancy between molecular body time and clock time may be tested as a marker of altered biological state, while accounting for sleep loss, inflammation, feeding, medication, and blood cell composition.
Recent computational approaches also support the need to move beyond single static markers. Wearable data can be integrated with mathematical modeling or data assimilation approaches to estimate circadian phase in real-world settings, but these methods also emphasize that wearable signals are indirect outputs of the circadian system rather than the clock itself [26]. Similar systems-modeling approaches have been proposed for personalized chronotherapy, where estimating individual internal time is necessary for adjusting treatment timing [27].
These forms of circadian misalignment are not yet validated clinical biomarkers in most settings. They are interpretive constructs. Their value is to organize hypotheses. If the main problem is phase-related misalignment, timed light or melatonin may be considered. If the main problem is weak or irregular zeitgeber exposure, daytime light and activity consolidation may be more relevant. If central-peripheral misalignment is suspected, meal timing, activity timing, and medication timing may need to be considered together with sleep-wake and light schedules. If trait-state misalignment is suspected, intervention may require more sustained rhythm stabilization. The next section discusses how this interpretation may guide individualized circadian intervention.
TOWARD TRAIT-STATE GUIDED CIRCADIAN INTERVENTION
The clinical goal of multilayered circadian assessment is not to assign patients to a fixed circadian category. A more realistic goal is to generate individualized intervention hypotheses. If several measures suggest phase-related misalignment, the intervention may focus on phase shifting. If the main problem is weak zeitgeber exposure, the intervention may focus on strengthening daytime cues and reducing inappropriate nighttime cues. If trait-state misalignment is suspected, intervention may need to be more sustained and individualized.
The most established intervention targets are light exposure and melatonin timing. Timed light and strategically timed melatonin are recommended for selected circadian rhythm sleep-wake disorders, although the strength of evidence differs by disorder and age group [28]. In this context, wearable-derived rhythm state may help identify when the patient is exposed to light, when activity occurs, and whether the sleep-wake schedule is stable. However, the timing of light or melatonin should ideally be interpreted in relation to internal circadian phase, because the same exposure can have different effects depending on biological time.
Sleep-wake scheduling is another practical target. A fixed wake time, regular sleep opportunity, and reduced evening light exposure may help stabilize rhythms in patients with delayed or irregular patterns. In patients with cognitive impairment or reduced daily structure, the goal may be less about shifting the clock and more about increasing daytime activity, strengthening daytime light exposure, and reducing rhythm fragmentation. This approach is consistent with clinical trials and reviews suggesting that light interventions may improve some sleep-related outcomes in dementia, while also showing heterogeneity across protocols and outcomes [28,29].
Meal timing and activity timing may be useful when central-peripheral misalignment is suspected. Human laboratory data show that delayed meal timing can shift some peripheral metabolic rhythms without necessarily shifting all markers of the central clock [30]. Exercise can also act as a nonphotic zeitgeber, and human phase-response data suggest that the direction and magnitude of phase shifting depend on exercise timing [31]. These interventions should not be presented as universal prescriptions. They are better understood as modifiable timing cues that may be adjusted according to the patient’s rhythm profile.
Social rhythm stabilization is also relevant. The clinical value of stabilizing daily routines has been developed most clearly in interpersonal and social rhythm therapy for bipolar disorder. This literature is not directly about neurodegenerative disease, but it provides a useful clinical precedent: regularity of sleep, meals, activity, and social contact can be treated as a therapeutic target rather than only as background lifestyle information [32,33]. For older adults or cognitively impaired patients, this principle may need to be adapted to include caregiver routines, daytime exposure to structured activity, and feasibility of adherence.
Timing of medication and other treatments is another possible extension. Some medications have time-dependent effects or adverse effects, and physiological targets such as blood pressure, sleepiness, cognition, and metabolism show circadian variation. Systems-modeling approaches have also been proposed for personalized chronotherapy, where estimating internal time is important for treatment timing [27]. However, medication timing should be treated conservatively in this review. It is a potential extension of circadian assessment, not the primary clinical recommendation.
In practice, the intervention target should depend on the most plausible form of circadian misalignment. Delayed internal phase may point to timed light or melatonin. Weak daytime zeitgeber exposure may point to daytime light, activity consolidation, and regular wake time. Central-peripheral misalignment may require attention to meal timing and activity timing in addition to sleep-wake and light schedules. Fragmented wearable-derived rhythm state may require rhythm stabilization rather than simple phase shifting. In neurodegenerative disease, this distinction may be important because rhythm disruption can reflect both biological vulnerability and reduced environmental structure.
Trait-state guided circadian intervention is not a validated clinical algorithm. It is a way to move from measurement to an intervention hypothesis. The goal is to interpret wearable-derived rhythm state, internal phase, patient-derived cellular period, and molecular timing information together, and then select modifiable timing cues such as light, sleep-wake schedule, activity, meals, social rhythm, and medication timing. In this sense, it represents a personalized approach to circadian medicine and may be considered one component of precision medicine.
CONCLUSION
Circadian information obtained from patients should not be treated as a single type of measurement. Wearable-derived rhythms, DLMO-based phase markers, patient-derived cellular period, and blood-based molecular timing provide complementary but noninterchangeable information. Each measure should be interpreted according to what it primarily reflects.
The value of combining these measures is to better understand circadian misalignment within the patient. Delayed sleep timing, fragmented rest-activity rhythm, altered cellular period, and shifted molecular body time may reflect different mechanisms, but they may also influence one another through light exposure, sleep-wake behavior, feeding, activity, inflammation, metabolism, and disease progression. In this sense, circadian assessment should move toward a systems-level interpretation of patient-specific temporal biology.
Trait-state guided circadian intervention should be viewed as a framework for moving from measurement to individualized intervention hypotheses. Future studies should test whether these measures are reproducible, stable over time, associated with clinical outcomes, and responsive to intervention. Such work may help develop a personalized and precision-medicine approach in which environmental timing, behavioral timing, and treatment timing are adjusted according to the patient’s measured circadian characteristics.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
