Bright light exposure (BLE) consists in using lamps for light therapy during 12 weeks, at the rate of five one-hour at-home sessions a week.

The GVS program lasts 2 weeks and consists in five 20-minutes at-home sessions a week (total of 10 sessions), using a portable galvanic stimulator.

The HE program lasts 14 weeks and consists in one one-hour at-home session a week (total of 14 sessions). This web-based intervention includes courses on general health (sleep habits, diet, physical activity, cardiovascular diseases, etc.).

Objective measure of sleep quality. The variable use to determine sleep quality is the sleep fragmentation index (SFI).

SFI is defined as the total number of awakenings and any sleep stage shift divided by the total sleep time ; and expressed in number of events per hour.

Objective measure of sleep quantity, expressed in minutes and based on bedtime and wake time.

Objective measure of sleep quantity expressed in minutes.

Objective measure of sleep quantity expressed in minutes.

Objective measure of sleep quantity expressed in minutes.

Objective measure of sleep quantity expressed in % and defined as 100 * total sleep time / time in bed.

Objective measure of sleep quality. Actigraphic measure of sleep quality is the fragmentation index defined as the sum of the Mobile time (%) and the Immobile bouts <=1min (%).

Part A of the State-Trait Anxiety Inventory (STAI-A ; Spielberger et al., 1970).

It consists in 20 statements for which volunteers have to evaluate how they feel at a particular moment in time.

Each answer is scored from 1 to 4, with 1 = low degree of anxiety, and 4 = high degree of anxiety.

Global score varies from 20 to 80, and means :

• from 20 to 35 : very low state anxiety
• from 36 to 45 : low state anxiety
• from 46 to 55 : medium state anxiety
• from 56 to 65 : high state anxiety
• from 66 to 80 : very high state anxiety

Volunteers are asked to place a vertical mark on a 100-mm horizontal line, between "calm" (on the left edge) and "anxious" (on the right edge) (Abend et al., 2014).

The Geriatric Depression Scale has been designed to specifically rate depression in older adults (Lach et al., 2010; Yesavage et al., 1982).

This scale includes 30 "yes / no" questions.

Global score varies between 0 and 30, and means :

• from 0 to 9 : no depression
• from 10 to 20 : minor depression
• from 21 to 30 : major depression

Evaluation of selective attention. Stroop test has three parts:

1. - Denomination condition: Volunteers have to name, as fast as possible, the color of rectangles printed in the card A.
2. - Reading condition: Volunteers have to read, as fast as possible, words printed in black and white in the card B.
3. - Interference condition: Volunteers have to name, as fast as possible, the color of the words printed in colors in the card C.

Dependant variables recorded in each condition:

• Time to perform the task
• Number of errors = number of self-corrected errors + non-self-corrected errors

Calculation of interference scores:

• Time in the Denomination condition - Time in the Interference condition
• Number of errors in the Denomination condition - Number of errors in the Interference condition

Two parts:

1. TMT-A: Processing speed: volunteers have to link, as fast as possible, numbers from 1 to 25 in sequential order.
2. TMT-B: Flexibility : volunteers have to alternate, as fast as possible, between numbers from 1 to 13 and letters from A to L : 1-A-2-B-3-C, etc.

Dependant variables recorded in each part:

• Time to perform the task
• Number of errors for TMT-A = number of self-corrected errors + non-self-corrected errors
• Number of errors for TMT-B = number of self-corrected errors + non-self-corrected errors + perseverative errors

Scoring procedure:

• Time to perform TMT-B - Time to perform TMT-A
• Number of errors for TMT-B - Number of errors for TMT-A

Assessment of working memory. First: determination of memory span. Single task 1: memory span for 2 minutes Single task 2: visuo spatial tracking task for 2 minutes Dual-task: both tasks are completed simultaneously for 2 minutes

Dependant variables recorded:

• Single task 1: number of correctly recalled series of items
• Single task 2: number of ticked boxes

Scoring procedure:

• Recall performance = (number of correctly recalled series of items during single task - number of correctly recalled series of items during dual-task) / number of correctly recalled series of items during single task.
• Tracking performance = (number of ticked boxes during single task - number of ticked boxes during dual task) / number of ticked boxes during single task.
• Mu index (index reflecting the performance decrement due to the dual-task) = (1-((Recall performance + Tracking performance)/2))*100. A high Mu index means high dual-task performance.

Computerized memory task, based on the Memory game (Rasch et al., 2007) comprising 24 cards (12 pairs of similar cards).

Learning condition: Volunteers have to find pairs of identical cards. Learning is completed when volunteers reach at least 68% of correct answers, i.e. when 16 cards are correctly assembled on 24.

Recall condition takes place the next morning. Only one trial is performed.

Dependant variables recorded:

• Number of pairs correctly recalled during the learning condition
• Number of pairs correctly recalled the next morning

Scoring procedure:

- Delta index: (number of pairs correctly recalled the next morning - number of pairs correctly recalled during the learning condition) / number of pairs correctly recalled during the learning condition

The hand grip strength is measured with a dynamometer. Volunteers complete three trials per hand, during 3 to 5 secondes, with 1-minute break between each trial to minimize fatigue effects.

Among the six measures, maximum values of each hand are retained (expressed in kg).

Leg extension power is assessed with the Nottingham power rig. Seated volunteers have to extend their tested leg (right leg) as fast and hard as possible by pushing unilaterally the footplate. There are eight trials, with 30-second break in between.

Dependant variables recorded:

• maximal leg power (expressed in watts)
• maximal leg power relative to the weight of the individual (expressed in watts/kg)

Postural balance is assessed through the Synapsys posturography system (Synapsys Inc.).

Volunteers stand on the Synapsys posturography system, feet apart, with their arms alongside their body, looking straight ahead.

Postural balance is measured in two conditions, with eyes open and eyes closed. There are two 30-second trials per condition.

For each trial, the position of the center of pressure (CoP) is recorded. Limit of stability test is recorded without time limitation. The maximum amplitude of oscillations of the CoP in mediolateral and anterioposterior directions are recorded.

The maximum incremental cardiopulmonary test is performed on an ergometric bicycle, based on an individualized protocol.

Prior to the test, a resting ECG is performed and verified by a cardiologist. Following a 3 min warm-up (30 watts), the power is incremented by 20 watts every 2 minutes until exhaustion. The test is followed by a 3-minute recovery period.

Maximum effort is determined by the occurrence of the following two criteria:

• 1) Presence of a VO2 plateau despite an increase of power (Åstrand et al., 1994);
• 2) Increase in oxygen consumption of the last increment should not exceed 100 ml/min despite the increase in power (Di Prampero et al., 1993).

Dependant variables recorded:

• Maximum oxygen consumption (VO2 max)
• Maximum aerobic power
• Maximum heart rate
• Respiratory quotient (RQ) at the end of test (RQ = Volume of CO2 / Volume of O2)

Questionnaire developed by the World Health Organization, which consists in 19 questions grouped to capture physical activity undertaken in different behavioural domains, as work, transport and discretionary (also known as leisure or recreation) (Cleland and al., 2014).

Within the work and discretionary domains, questions assess the frequency and duration of two different categories of activity defined by the energy requirement or intensity (vigorous or moderate intensity).

In the transport domain, the frequency and duration of all walking and cycling for transport is captured but no attempt is made to differentiate between these activities.

One additional item collects time spent in sedentary activities.

Calculation of a global MET score (Metabolic Equivalent of Task, 1 MET = 1 kcal/kg/hour).

Higher MET score means high physical activity.

Assessment of dynamic balance in the older adults (Mathias et al., 1986; Podsiadlo & Richardson, 1991).

Volunteers are asked to sit down on a chair with their feet on the floor and their hands on their legs. At the "go" signal, they have to stand up, walk 3 meters, turn around a cone, walk back, and sit down again.

The performance time starts when volunteers start moving to stand up and ends when they sit down again properly.

The test is performed in two conditions:

• Two trials at a self-paced velocity, and
• Two trials as fast as possible

Duration of the two best trials of the two conditions are retained (Podsiadlo & Richardson, 1991).