Melatonin is a safe and accessible sleep-promoting supplement. The tablet will be taken once daily, 30 minutes before habitual bedtime for 4 weeks, concurrent with CBT-I.

Four group-based weekly sessions covering sleep hygiene, healthy diet, and exercise habits for older adults. This arm will not include any active therapeutic CBT-I components. It serves as the active control, controlling for attention and non-specific expectancy effects.

Placebo capsule identical in appearance to the melatonin capsule, taken once daily, 30 minutes before habitual bedtime for 4 weeks. Used in CBT-I + Placebo and Psychoeducation + Placebo arms.

Pittsburgh Sleep Quality Index (PSQI) is a self-rated scale consisting of 19 questions. All items are combined to form seven component scores on different aspects of sleep quality, each of which ranges from 0 to 3 points with higher scores representing more sleep disturbance. The seven component scores are added to one global score, which ranges from 0 to 21, with higher scores indicating more difficulties with sleep.

The Munich Chronotype Questionnaire (MCTQ) is a self-report measures of sleeping patterns during weekdays and weekends separately. The Mid-Sleep Time (MSF/MSFsc) are used to as an indicator of chronotype, where individuals with earlier mid-sleep time reflect a morning chronotype and later mid-sleep time reflect an evening chronotype.

The Morningness-Eveningness Questionnaire (MEQ) is a 19-item self-assessment tool used to determine an individual's chronotype with a total score between 16 and 86. A higher score indicates a morning chronotype (e.g., "morning type"), a lower score indicates an evening chronotype (e.g., "evening type"), and a mid-range score indicates an intermediate chronotype.

Hospital Anxiety and Depression Scale (HADS), a 14-item self-report questionnaire used to screen for anxiety and depression in a patient's recent past week. The depression subscale ranges in scores from 0 to 21, with higher scores indicating more severe states of depression. Similarly, the anxiety subscale ranges in scores from 0-21 with higher scores indicating more severe states of anxiety.

Epworth sleepiness scale, a short self-assessment to identify how likely you are to fall asleep during the daytime. Scores range from 0 to 24, a higher score indicates a higher level of excessive daytime sleepiness.

Multidimensional Fatigue Inventory (MFI) is a 20-item self-rated scale on fatigue symptoms. There are three subscales, measuring the physical (possibly scored from 7 to 35), mental (possibly scored from 6 to 30), and spiritual (possibly scored from 7 to 35), dimensions of fatigue. A grand total score can be calculated by summing up the three sub scores. In all cases, a higher score represents higher fatigue symptoms.

Clinical Global Impression (CGI) Scale is a clinician-rated scale, comprised of two one-item subscales: Severity of Illness (CGI-S) subscale evaluating the severity of psychopathology, and Clinical Global Improvement Scale (CGI-I) evaluating change from the initiation of treatment. In both cases, the score is given on a seven-point scale, with higher values indicating higher severity of illness and larger improvement, respectively. CGI-S will be used to evaluate the overall severity of clinical symptoms.

Clinical Global Impression (CGI) Scale is a clinician-rated scale, comprised of two one-item subscales: Severity of Illness (CGI-S) subscale evaluating the severity of psychopathology, and Clinical Global Improvement Scale (CGI-I) evaluating change from the initiation of treatment. In both cases, the score is given on a seven-point scale, with higher values indicating higher severity of illness and larger improvement respectively. Overall treatment response is defined as CGI-I <3 (clinician-rated)

The SF-36, or Short Form 36 Health Survey, is a 36-item patient-reported questionnaire that measures health status and quality of life. It assesses eight different health domains, including physical functioning, bodily pain, vitality, general health, role-physical, role-emotional, social functioning, and mental health. Total scores range from 0 to 100, where higher scores indicate better health.

The stop-signal task is a behavioural task that measures a person's ability to inhibit a pre-programmed response. The main measure of inhibitory control is the stop-signal reaction time (SSRT), which estimates the time it takes to cancel a response. A longer SSRT indicates a slower ability to inhibit a response.

Psychomotor Vigilance Task (PVT) is a reaction-time test that measures sustained attention by having a person respond as quickly as possible to a visual stimulus that appears at random intervals. Reaction Time (RT): The primary measurement is how quickly the participant responds to the stimulus. Lapses: A lapse is a failure to respond to the stimulus within a certain time frame, often defined as a reaction time longer than 500 milliseconds.

N-back Task for assessing working memory capacity and manipulation. In N-back Task, a d prime score will be calculated based on the signal detection theory, where a higher score indicates better working memory performance.

Wisconsin Card Sorting Test for assessing problem solving. In Wisconsin Card Sorting Test, lower executive functioning is indicated by a higher percentage of persistent errors and a higher number of trials taken to complete the first category.

Chinese Auditory Verbal Learning Task for assessing episodic memory, where a higher number of recalled words indicates better episodic memory performance.

Multifactorial Memory Questionnaire (MMQ) is a 57-item self-report assessing memory satisfaction, ability and strategy use (three subscales).For each subscale, score ranges: for each subscale: Satisfaction (0-72), Ability (0-80), and Strategy (0-76). A higher score generally indicates better self-reported memory functioning, higher satisfaction with memory, and greater use of memory strategie

Beck Depression Inventory-II (BDI-II) is 21-item self-report of depressive-symptom severity during the past two weeks (0-63). Higher scores indicate more severe depressive symptoms

The Hamilton Rating Scale for Depression (HAM-D17) is a 17-item interview to assess the severity of depression symptoms. Scores are interpreted based on a range: 0-7 indicates no depression, 8-16 is mild, 17-23 is moderate, and 24 or higher is severe depression.

Dysfunctional Beliefs and Attitudes about Sleep Scale (DBAS): 16 statements evaluating maladaptive sleep-related cognitions (0-160). A higher score indicates a higher level of dysfunctional, unhelpful, or rigid beliefs about sleep.

Cognitive Failures Questionnaire; 25 items indexing self-reported failures in attention, memory and action execution, 0-100. Indicates a higher frequency of everyday cognitive failures.

Pre-Sleep Arousal Scale (PSAS): 16 items measuring somatic and cognitive arousal at bedtime (8 each; 1-5 Likert, total score range 8-40). Higher scores indicate greater pre-sleep physical distress.

Ford Insomnia Response to Stress Test (FIRST): 9-item trait measure of likelihood that stress will provoke insomnia (0-36).

Sleep Hygiene Practices Scale (SHPS): 30 items assessing frequency of behaviours affecting sleep quality (0-120). Higher scores indicate more maladaptive (unhealthier) sleep hygiene practices

Daily sleep diary for consecutive seven days. Sleep parameter estimated by daily sleep diary: time in bed (TIB) in hours

Daily sleep diary for consecutive seven days. Sleep parameter estimated by daily sleep diary: total sleep time (TST) in hours

Daily sleep diary for consecutive seven days. Sleep parameter estimated by daily sleep diary: sleep onset latency (SOL) in mins

Daily sleep diary for consecutive seven days. Sleep parameter estimated by daily sleep diary: wake after sleep onset (WASO) in mins

Daily sleep diary for consecutive seven days. Sleep parameter estimated by daily sleep diary: sleep efficiency (SE), which is calculated by total sleep time divided by total time in bed, %

The duration of time spent in bed attempting to sleep, estimated from 7-night wrist actigraphy. Reported as mean minutes per night across the recording period.

The total amount of sleep obtained during a sleep period, estimated from 7-night wrist actigraphy. Reported as mean minutes per night across the recording period.

The amount of time it takes to fall asleep after attempting to do so, estimated from 7-night wrist actigraphy. Reported as mean minutes per night across the recording period.

The amount of time spent awake after initially falling asleep, estimated from 7-night wrist actigraphy. Reported as mean minutes per night across the recording period.

The standard deviation of the sleep midpoint, defined as (sleep onset + sleep offset) / 2, across the 7-night recording. Higher values reflect more irregular sleep timing across nights.

The percentage of time spent asleep while in bed, calculated as (TST / TIB) × 100, estimated from 7-night wrist actigraphy.

The clock time of maximum activity estimated from a cosinor model fitted to the 7-day rest-activity record. Represents the timing of peak daily activity.

The difference between the peak and nadir of the cosinor-fitted rest-activity curve across the 7-day recording. Reflects the overall strength of the circadian rest-activity rhythm.

An index ranging from 0 to 2 quantifying the degree of rest-activity rhythm fragmentation. Higher values reflect greater fragmentation, indicating more daytime napping or nocturnal awakening.

An index ranging from 0 to 1 reflecting the synchronisation of the rest-activity rhythm to the 24-hour light-dark cycle. Higher values reflect stronger synchronisation.

The clock time of onset of the 5 least active consecutive hours across the 7-day recording, presumed to represent the habitual sleep window.

Actigraphic assessment for consecutive seven days. Circadian parameters computed by the nonparametric circadian rhythm analysis method - start times and average activity of M10 (i.e. 10 h with maximal activity).

Actigraphic assessment for seven consecutive days. The standard deviation of the midpoint of sleep will be derived from actigraphy, defined as sleep onset + sleep offset) / 2, across the 7-night recording.

DLMO time to sleep-onset phase angle: Interval (min) from DLMO time to diary-determined sleep-onset time. Sleep-onset time will be calculated as the mean sleep-onset time across valid diary nights.

Dim-light melatonin onset (express as time value hh:mm) is determined by 6-hours salivary melatonin collected at 30-minutes interval. Melatonin is assayed by gas chromatography-mass spectrometry.

DLMO time: Clock time (hh:mm) at which salivary melatonin first exceeds the 3 pg mL-¹ threshold under < 10 lux illumination.

DLMO time to sleep-midpoint phase angle is the initerval (min) from DLMO time to diary-determined sleep midpoint. Sleep midpoint will be calculated as halfway between sleep onset and final sleep offset/waketime.

Interval (min) from DLMO time to diary-determined lights-off/bedtime; indexing the timing of attempted sleep relative to biological night onset. Lights-off/bedtime will be calculated as the mean diary-reported lights-off/bedtime across valid diary nights.

T75 recovery time is the seconds from stimulus offset to 75% return toward baseline pupil diameter, reflecting the post-illumination pupil response (PIPR), a melanopsin-mediated circadian biomarker.

Objective sleep parameter derived from overnight polysomnography: time in bed (TIB) in hours, measuring the total duration from lights-out to lights-on.

Objective sleep parameter derived from overnight polysomnography: total sleep time (TST) in hours.

Objective sleep parameter derived from overnight polysomnography: sleep efficiency (SE) expressed as a percentage, calculated as total sleep time divided by time in bed.

Objective sleep parameter derived from overnight polysomnography: sleep latency in minutes, measuring the time interval from lights-off to the onset of the first scored sleep stage.

Objective sleep parameter derived from overnight polysomnography: wake after sleep onset (WASO) in minutes, measuring the total duration of wakefulness recorded after initial sleep onset.