To assess functioning in the Behavioral domain, a Youth Self-Report Composite Risk Score is calculated by taking the mean of standardized summary scores from two measures: Sensation Seeking Scale for Children (SSS) and the Alcohol and Substance Use-Past 30 days (SU; items include questions on caffeine and energy drinks).

Summary scores were calculated as 1.) SSS = sum of 8 items, rated 1-5, range: 8-40. Higher scores = greater sensation seeking; and 2.) SU = Sum of 23 items rated 0-7, range 0-161. Higher scores = more frequent use.

For each participant, summary scores from the SSS and SU subscales were computed and standardized. The final composite (range -1.73 to 3.34) was calculated as the mean of the two standardized scores. Higher composite scores indicate higher impairment Change in this composite score from baseline to post-treatment is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Youth Self-Report Composite Score, Emotional functioning domain is calculated by taking the mean of standardized summary scores from two measures: Children's Depression Rating Scale-Revised (CDRS) and the Multidimensional Anxiety Scale for Children (MASC).

Summary scores were calculated as 1.) CDRS = Sum of 17 items. range 17-113. Higher scores indicate greater depressive symptoms; and 2.) MASC= Sum of 39 items. range 0-117. Higher scores indicate greater anxiety.

For each participant, summary scores from the CDRS and MASC subscales were first computed and then standardized. The final composite (range -2.10 to 3.60) was calculated as the mean of the two standardized scores from the CDRS and MASC. Higher composite scores indicate greater emotional risk (i.e., more depression and anxiety symptoms).

Change in this composite score from baseline to post treatment is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Score for Social Domain to assess functioning in the Social domain. Calculated by taking the average of the three subscales (ie, friends, family, romantic relationships) from the Youth Social Adjustment Scale - Self Report.

Youth Social Adjustment Scale - social items only. 9 items (questions 7-15 out of the 23 item scale), item range 1-5. Summary score is calculated by taking the sum of all 9 items. no reverse coding needed. Summary score range 9-45. Higher score = more impaired adjustment.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Score for Physical Domain (physical functioning) is calculated by taking the mean of standardized summary scores from two measures: Modifiable Activity Questionnaire for Adolescents (MAQ) and Physical Health Questionnaire (PHQ).

Summary scores were calculated as 1.) MAQ = sum of the number of hours per week not active/exercising. Higher scores indicate greater numbers of leisure hours; and 2. PHQ-15 = sum of item (13 items for males, 14 items for females), range 0-30. Higher scores indicate worse somatic complaints.

For each participant, summary scores from the MAQ and PHQ-15 subscales were first computed and then standardized. The final composite (range -3.95 to 1.86) was calculated as the mean of the two standardized scores. Higher composite scores indicate greater physical health risk.

Change in this composite score from baseline to post treatment is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Melatonin levels were measured using 13 saliva samples collected at 30-minute intervals, beginning 5.5 hours before and ending 30 minutes after each participant's average bedtime (computed from 7 nights of sleep diary). For each timepoint (baseline and post-treatment), the dim light melatonin onset (DLMO) was estimated by identifying when melatonin levels crossed the 3.0 pg/ml threshold (i.e., the interpolated time). Change in the interpolated DLMO times from baseline to post treatment (which is an average of 9 weeks after the beginning of treatment) reported. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Pittsburgh Sleep Quality Index (PSQI). Summary score is calculated by taking the sum of item-level scores. Summary score range 0-21. A higher score means increased severity of difficulty in all sleep area components. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

The discrepancy between weeknights and weekends for Total Sleep Time(TST). The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Emotional Health, measured via Ecological Momentary Assessment (EMA), assesses subjective emotional well-being across 7-days, per timepoint using a 9-item short form of the Positive and Negative Affect Scale for Children (PANAS-C), which included 4 positive affect items (min = 4, max = 20; higher scores indicate more positive affect) and 5 negative affect items (min = 5, max = 25; higher scores indicate more negative affect), each rated on a 5-point Likert scale.

A positivity ratio was calculated by dividing the sum of each affect item scores by the sum of negative affect score per survey day. Higher positivity ratios (range: 0.16-4) indicate higher subjective well-being and less risk.

Change in average Positivity Ratio from baseline to post treatment (which is an average of 9 weeks after the beginning of treatment) reported. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. The Emotional Health Composite Risk Score was assessed with a composite score of the Anxious/Depressed and Withdrawn/Depressed subscales of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's emotional health, including items related to anxiety/depression. Range for composite total: 0-42, with higher scores indicating more problems.

The CBCL Anxious/Depressed subscale is composed of 13-items CBCL items on a scale of 0-2, the range of scores is 0-26, with higher scores indicating more emotional problems.

The CBCL Withdrawn/Depressed subscale is composed of 8-items CBCL items on a scale of 0-2, the range of scores is 0-16, with higher scores indicating more emotional problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition

The discrepancy between weeknights and weekends for Bedtime(BT) via Daily Sleep Diary. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

The discrepancy between weeknights and weekends for Wake time (WUP) via Daily Sleep Diary. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Cognitive Health measured via Ecological Momentary Assessment. Participants responded to 11 items that measured their concentration, distractedness, and focus related to their current activity. All items were rated on 5-point Likert scale, where higher scores respectively indicated higher levels of concentration, distractedness, and focus (min = 11, max = 55). The composite risk score was calculated by averaging participant responses over the assessment week. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Behavioral Health measured via Ecological Momentary Assessment was assessed by directly asking participants about eating, drinking, chewing gum, and smoking behavior at the time the phone rang through 6 open-ended EMA questions. Responses were coded. The average weekly frequency of intake of junk food, caffeine, alcohol, nicotine, and other substances was tabulated. The minimum score was 0, and there was no maximum score. Higher scores indicate more risky behaviors and thus more risk. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Physical Health derived from Ecological Momentary Assessment (EMA). For the physical health domain, responses to a single daily item (e.g., "Were you physically active today?") were used to create a binary score: 1 = active, 2 = inactive. Daily responses were collected over 7 days and both summed (range: 7-14) and averaged for each participant.

The final composite score reflects the average of these daily values (range: 1 - 2) , where lower scores indicate greater physical activity and lower physical health risk.

Change in this composite score from baseline to post treatment (which is an average of 9 weeks after the beginning of treatment) is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Social health measured via Ecological Momentary Assessment was assessed by directly asking participants who the participant was with at the time the phone rang. Participants' responses were then manually coded for companion type on a scale of 0-4. Participants' positivity ratios, as calculated in the Emotional Health (Positivity Ratio) Composite Risk Score of Functioning, were then grouped based on occasions when participants were alone vs. with a family member vs. with a friend vs. other. Positivity ratios within groups were averaged (min = 0.16, max = 4), where higher averages indicated higher subjective well-being and lower risk. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. Cognitive Health Composite Risk Score was assessed with the Thought Problems and Attention Problems subscale of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's cognitive health, including items related to thought problems. Range for composite total: 0-50, with higher scores indicating more cognitive problems.

The CBCL Thought Problems subscale is composed of 15-items CBCL items on a scale of 0-2, the range of scores is 0-30, with higher scores indicating more problems.

The CBCL Attention Problems subscale is composed of 10-items CBCL items on a scale of 0-2, the range of scores is 0-20, with higher scores indicating more problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition

Parent Measure. The Behavioral Health Composite Risk Score was assessed with a composite score of the Rule-Breaking Behavior and Aggressive Behavior subscales of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's behavioral health, including items related to rule-breaking. Range for composite total: 0-70, with higher scores indicating more behavioral problems.

The CBCL Rule-Breaking Behavior subscale is composed of 17-items CBCL items on a scale of 0-2, the range of scores is 0-34, with higher scores indicating more rule-breaking behavior.

The CBCL Aggressive Behavior subscale is composed of 18-items CBCL items on a scale of 0-2, the range of scores is 0-36, with higher scores indicating more aggressive behavior.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. The Social Health Composite Risk Score was assessed with a composite score of the Social Problems subscale of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's social health. The CBCL Social Problems subscale is composed of 11-items CBCL items on a scale of 0-2, the range of scores is 0-22, with higher scores indicating more Social problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. The Physical Health Composite Risk Score was assessed with a composite score of the Somatic Complaints subscale of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's Physical health (e.g., "vomiting").

The CBCL Somatic Complaints subscale is composed of 11-items CBCL items on a scale of 0-2, the range of scores is 0-22, with higher scores indicating more physical problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. The Sleep Health Composite Risk Score was assessed with a composite score of sleep from Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's sleep health. The CBCL sleep subscale is composed of 7-items CBCL items on a scale of 0-2, the range of scores is 0-14, with higher scores indicating more sleep problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

calculated separately for weeknights and weekend nights to also compute the discrepancy between weeknights and weekend nights

Actigraphy Measured Sleep Onset Latency (SOL; calculated separately for weeknights and weekend nights) and Wake After Sleep Onset (WASO; calculated separately for weeknights and weekend nights) to create Total Wake Time (SOL+WASO) for weeknights, weekends and to also compute the discrepancy between weeknights and weekend nights

calculated separately for weekdays and weekends

total sleep time (weekend nights), weekend night bedtime, weekend rise time, total wake time (SOL+WASO) weeknights, total wake time (SOL+WASO) weekend nights, naps.

from females only

collected via saliva sample DNA by saliva sample to examine selected circadian and emotion single nucleotide polymorphism (SNP)

collected via saliva sample

Credibility Expectancy Questionnaire

Parent measure

Parent measure

Parent measure

Total of all reporting groups

Youth aged 10 to 18 years with an evening chronotype and who were"at risk"in 1 of 5 health domains.

Participants included 176 youth recruited through clinicians or advertisements.

K-SADS = Schedule for Affective Disorders and Schizophrenia for School-Age Children. The K-SADS was administered separately to youth and one parent/caregiver to assess for participant current and lifetime Axis I disorders at the pretreatment assessment.

The number analyzed includes only those with data on this variable.

K-SADS = Schedule for Affective Disorders and Schizophrenia for School-Age Children. The K-SADS was administered separately to youth and one parent/caregiver to assess for participant current and lifetime Axis I disorders at the pretreatment assessment.

The number analyzed includes only those with data on this variable.

K-SADS = Schedule for Affective Disorders and Schizophrenia for School-Age Children. The K-SADS was administered separately to youth and one parent/caregiver to assess for participant current and lifetime Axis I disorders at the pretreatment assessment.

The number analyzed includes only those with data on this variable.

K-SADS = Schedule for Affective Disorders and Schizophrenia for School-Age Children. The K-SADS was administered separately to youth and one parent/caregiver to assess for participant current and lifetime Axis I disorders at the pretreatment assessment.

The number analyzed includes only those with data on this variable.

Change in average bedtime on weeknights from pre-treatment to post-treatment measured via Daily Sleep Diary. 24-hour decimal format, where times after midnight are expressed as numbers above 24 (ex. 1:30 am is 25.50). The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Morning Eveningness preference measured via Childrens Morningness Eveningness Preference Scale(CMEP). Scores range from 10 (Extreme evening preference) to 43 (Extreme morning preference). The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Cognitive composite score reflects cognitive functioning and was calculated by averaging the standardized summary scores from two measures: Attentional Control Scale (ACS) and Youth Social Adjustment Scale-Self Report (YSAS). Summary scores were calculated as 1.)ACS = sum of 20 items rated 1- 4; range: 20-80. Higher scores indicate less attentional control; and 2.) YSAS = sum of 6 school/cognitive-related items rated 1-5; range: 6-30. Higher scores indicate worse school-related impairment.

Per participant, summary scores from the ACS and YSAS were computed and then standardized.

The final composite (range -2.12 to 2.39) was calculated as the mean of the two standardized scores. Higher scores indicate greater attentional difficulty and school impairment.

Change in this composite score from baseline to post treatment is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

To assess functioning in the Behavioral domain, a Youth Self-Report Composite Risk Score is calculated by taking the mean of standardized summary scores from two measures: Sensation Seeking Scale for Children (SSS) and the Alcohol and Substance Use-Past 30 days (SU; items include questions on caffeine and energy drinks).

Summary scores were calculated as 1.) SSS = sum of 8 items, rated 1-5, range: 8-40. Higher scores = greater sensation seeking; and 2.) SU = Sum of 23 items rated 0-7, range 0-161. Higher scores = more frequent use.

For each participant, summary scores from the SSS and SU subscales were computed and standardized. The final composite (range -1.73 to 3.34) was calculated as the mean of the two standardized scores. Higher composite scores indicate higher impairment Change in this composite score from baseline to post-treatment is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Youth Self-Report Composite Score, Emotional functioning domain is calculated by taking the mean of standardized summary scores from two measures: Children's Depression Rating Scale-Revised (CDRS) and the Multidimensional Anxiety Scale for Children (MASC).

Summary scores were calculated as 1.) CDRS = Sum of 17 items. range 17-113. Higher scores indicate greater depressive symptoms; and 2.) MASC= Sum of 39 items. range 0-117. Higher scores indicate greater anxiety.

For each participant, summary scores from the CDRS and MASC subscales were first computed and then standardized. The final composite (range -2.10 to 3.60) was calculated as the mean of the two standardized scores from the CDRS and MASC. Higher composite scores indicate greater emotional risk (i.e., more depression and anxiety symptoms).

Change in this composite score from baseline to post treatment is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Score for Social Domain to assess functioning in the Social domain. Calculated by taking the average of the three subscales (ie, friends, family, romantic relationships) from the Youth Social Adjustment Scale - Self Report.

Youth Social Adjustment Scale - social items only. 9 items (questions 7-15 out of the 23 item scale), item range 1-5. Summary score is calculated by taking the sum of all 9 items. no reverse coding needed. Summary score range 9-45. Higher score = more impaired adjustment.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Score for Physical Domain (physical functioning) is calculated by taking the mean of standardized summary scores from two measures: Modifiable Activity Questionnaire for Adolescents (MAQ) and Physical Health Questionnaire (PHQ).

Summary scores were calculated as 1.) MAQ = sum of the number of hours per week not active/exercising. Higher scores indicate greater numbers of leisure hours; and 2. PHQ-15 = sum of item (13 items for males, 14 items for females), range 0-30. Higher scores indicate worse somatic complaints.

For each participant, summary scores from the MAQ and PHQ-15 subscales were first computed and then standardized. The final composite (range -3.95 to 1.86) was calculated as the mean of the two standardized scores. Higher composite scores indicate greater physical health risk.

Change in this composite score from baseline to post treatment is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Sleepiness subscale from School Sleep Habits Survey (SSHB). This subscale includes 10 items rated on a 4-point scale (0 - 3), assessing sleepiness. Total scores were calculated by summing all item responses. Possible scores range from 0 to 30, with higher scores indicating greater daytime sleepiness. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Melatonin levels were measured using 13 saliva samples collected at 30-minute intervals, beginning 5.5 hours before and ending 30 minutes after each participant's average bedtime (computed from 7 nights of sleep diary). For each timepoint (baseline and post-treatment), the dim light melatonin onset (DLMO) was estimated by identifying when melatonin levels crossed the 3.0 pg/ml threshold (i.e., the interpolated time). Change in the interpolated DLMO times from baseline to post treatment (which is an average of 9 weeks after the beginning of treatment) reported. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Pittsburgh Sleep Quality Index (PSQI). Summary score is calculated by taking the sum of item-level scores. Summary score range 0-21. A higher score means increased severity of difficulty in all sleep area components. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

The discrepancy between weeknights and weekends for Total Sleep Time(TST). The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Emotional Health, measured via Ecological Momentary Assessment (EMA), assesses subjective emotional well-being across 7-days, per timepoint using a 9-item short form of the Positive and Negative Affect Scale for Children (PANAS-C), which included 4 positive affect items (min = 4, max = 20; higher scores indicate more positive affect) and 5 negative affect items (min = 5, max = 25; higher scores indicate more negative affect), each rated on a 5-point Likert scale.

A positivity ratio was calculated by dividing the sum of each affect item scores by the sum of negative affect score per survey day. Higher positivity ratios (range: 0.16-4) indicate higher subjective well-being and less risk.

Change in average Positivity Ratio from baseline to post treatment (which is an average of 9 weeks after the beginning of treatment) reported. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. The Emotional Health Composite Risk Score was assessed with a composite score of the Anxious/Depressed and Withdrawn/Depressed subscales of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's emotional health, including items related to anxiety/depression. Range for composite total: 0-42, with higher scores indicating more problems.

The CBCL Anxious/Depressed subscale is composed of 13-items CBCL items on a scale of 0-2, the range of scores is 0-26, with higher scores indicating more emotional problems.

The CBCL Withdrawn/Depressed subscale is composed of 8-items CBCL items on a scale of 0-2, the range of scores is 0-16, with higher scores indicating more emotional problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition

The discrepancy between weeknights and weekends for Bedtime(BT) via Daily Sleep Diary. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

The discrepancy between weeknights and weekends for Wake time (WUP) via Daily Sleep Diary. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Cognitive Health measured via Ecological Momentary Assessment. Participants responded to 11 items that measured their concentration, distractedness, and focus related to their current activity. All items were rated on 5-point Likert scale, where higher scores respectively indicated higher levels of concentration, distractedness, and focus (min = 11, max = 55). The composite risk score was calculated by averaging participant responses over the assessment week. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Behavioral Health measured via Ecological Momentary Assessment was assessed by directly asking participants about eating, drinking, chewing gum, and smoking behavior at the time the phone rang through 6 open-ended EMA questions. Responses were coded. The average weekly frequency of intake of junk food, caffeine, alcohol, nicotine, and other substances was tabulated. The minimum score was 0, and there was no maximum score. Higher scores indicate more risky behaviors and thus more risk. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Physical Health derived from Ecological Momentary Assessment (EMA). For the physical health domain, responses to a single daily item (e.g., "Were you physically active today?") were used to create a binary score: 1 = active, 2 = inactive. Daily responses were collected over 7 days and both summed (range: 7-14) and averaged for each participant.

The final composite score reflects the average of these daily values (range: 1 - 2) , where lower scores indicate greater physical activity and lower physical health risk.

Change in this composite score from baseline to post treatment (which is an average of 9 weeks after the beginning of treatment) is reported below. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Composite Risk Score of Functioning in Social health measured via Ecological Momentary Assessment was assessed by directly asking participants who the participant was with at the time the phone rang. Participants' responses were then manually coded for companion type on a scale of 0-4. Participants' positivity ratios, as calculated in the Emotional Health (Positivity Ratio) Composite Risk Score of Functioning, were then grouped based on occasions when participants were alone vs. with a family member vs. with a friend vs. other. Positivity ratios within groups were averaged (min = 0.16, max = 4), where higher averages indicated higher subjective well-being and lower risk. The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. Cognitive Health Composite Risk Score was assessed with the Thought Problems and Attention Problems subscale of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's cognitive health, including items related to thought problems. Range for composite total: 0-50, with higher scores indicating more cognitive problems.

The CBCL Thought Problems subscale is composed of 15-items CBCL items on a scale of 0-2, the range of scores is 0-30, with higher scores indicating more problems.

The CBCL Attention Problems subscale is composed of 10-items CBCL items on a scale of 0-2, the range of scores is 0-20, with higher scores indicating more problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition

Parent Measure. The Behavioral Health Composite Risk Score was assessed with a composite score of the Rule-Breaking Behavior and Aggressive Behavior subscales of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's behavioral health, including items related to rule-breaking. Range for composite total: 0-70, with higher scores indicating more behavioral problems.

The CBCL Rule-Breaking Behavior subscale is composed of 17-items CBCL items on a scale of 0-2, the range of scores is 0-34, with higher scores indicating more rule-breaking behavior.

The CBCL Aggressive Behavior subscale is composed of 18-items CBCL items on a scale of 0-2, the range of scores is 0-36, with higher scores indicating more aggressive behavior.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. The Social Health Composite Risk Score was assessed with a composite score of the Social Problems subscale of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's social health. The CBCL Social Problems subscale is composed of 11-items CBCL items on a scale of 0-2, the range of scores is 0-22, with higher scores indicating more Social problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. The Physical Health Composite Risk Score was assessed with a composite score of the Somatic Complaints subscale of Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's Physical health (e.g., "vomiting").

The CBCL Somatic Complaints subscale is composed of 11-items CBCL items on a scale of 0-2, the range of scores is 0-22, with higher scores indicating more physical problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Parent Measure. The Sleep Health Composite Risk Score was assessed with a composite score of sleep from Child Behavior Checklist (CBCL), calculated by taking the mean of standardized summary scores for the items within this domain. This is a parent-reported measure of a teen's sleep health. The CBCL sleep subscale is composed of 7-items CBCL items on a scale of 0-2, the range of scores is 0-14, with higher scores indicating more sleep problems.

The model provides estimates of the mean pre-post change in the PE condition and the TranS-C condition.

Measured via Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children (KSADS) Interview

Measured via Duke Structured Interview for Sleep Disorders

calculated separately for weeknights and weekend nights to also compute the discrepancy between weeknights and weekend nights

Actigraphy Measured Sleep Onset Latency (SOL; calculated separately for weeknights and weekend nights) and Wake After Sleep Onset (WASO; calculated separately for weeknights and weekend nights) to create Total Wake Time (SOL+WASO) for weeknights, weekends and to also compute the discrepancy between weeknights and weekend nights

calculated separately for weekdays and weekends

total sleep time (weekend nights), weekend night bedtime, weekend rise time, total wake time (SOL+WASO) weeknights, total wake time (SOL+WASO) weekend nights, naps.