Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jan 2.
Published in final edited form as: Behav Sleep Med. 2024 Oct 6;23(1):92–104. doi: 10.1080/15402002.2024.2412331

What Keeps Night Owls Well During the Week? Sleep Onset Consistency as a Moderator Between Morningness–Eveningness and Adolescent Development

Brian T Gillis 1, Mina Shimizu 1, Emily F Brigham 1, Ben Hinnant 1, Mona El-Sheikh 1
PMCID: PMC11693466  NIHMSID: NIHMS2027246  PMID: 39370647

Abstract

Objectives:

A preference for eveningness – one’s perception of being most alert later in the day – is associated with negative developmental outcomes in adolescence. Sleep onset consistency is protective against such outcomes. Toward a more nuanced understanding of relations between sleep-wake processes and adolescent development, we examined weeknight sleep onset consistency as a moderator of relations between eveningness and multiple indicators of development.

Method:

A sample of 272 high-school students (Mage = 17 years, SD = 9.12 months; n = 133 identified as female; 41% non-Hispanic Black/African-American, 59% non-Hispanic White/European-American) participated in a week of at-home sleep actigraphy assessment in 2017–2018. Adolescents reported their morningness – eveningness, internalizing symptoms (depression, anxiety), positive affect (optimism and subjective happiness), and physical health, and mothers reported on youths’ behavior problems. Relations were examined between morningness – eveningness and each indicator of development; sleep onset consistency was examined as a moderator of these associations.

Results:

On average, adolescents with a preference for eveningness had higher levels of externalizing behaviors and internalizing symptoms and lower levels of positive affect and physical health compared to peers with a preference for morningness (Bs = –0.27*–0.12***). Each association was moderated by weeknight sleep onset consistency. Across all indicators of development, evening-preferring youth with more consistent weeknight sleep onset had 0.49–0.72 SD better outcomes on average than evening-preferring youth with less consistent weeknight sleep onset.

Conclusions:

Falling asleep at roughly the same time each night can protect adolescent night owls from behavior problems and internalizing symptoms and can promote their positive affect and physical health.

Across the lifespan, the location of the primary sleep phase within a 24-hr period varies among individuals (Panjeh et al., 2021). Such variations occur along two distinct dimensions: (1) morningness – eveningness (M – E) – the sense of being more alert at earlier or later times of day (i.e., circadian preference) and (2) sleep schedule – the chronological timing of sleep (Adan et al., 2012). Having a preference for eveningness (colloquially referred to as being a “night owl”) and going to sleep at later times are associated with increased risk for behavior and emotional problems, diminished positive affect, and physical illness (Cheung et al., 2023; Cooper et al., 2023; Jankowski & Linke, 2020). Consistency in sleep onset – falling asleep around the same time each night – holds promise to protect night owls because this key sleep dimension is associated with fewer behavior and emotional problems and better physical health (Becker et al., 2017). The present study examines sleep onset consistency as a moderator of relations between M – E and four dimensions of adolescent development: externalizing (behavior) problems, internalizing symptoms (depression/anxiety), positive affect (optimism/happiness), and physical health.

Morningness–eveningness and adolescent development

An ample literature provides evidence of relations between M – E and adolescent development. A preference for eveningness in adolescence is associated with externalizing problems, including aggressive, rule-breaking, and antisocial behavior (Cooper et al., 2023; Tosuntaş et al., 2020), as well as internalizing symptoms of depression(S.-J. Chen et al., 2021) and anxiety (Tosuntaş et al., 2020). In contrast, a preference for morningness (i.e., less eveningness) is linked to lower levels of internalizing symptoms (Azad Marzabadi & Amiri, 2017). Other studies have found associations between M – E preferences and positive affect, demonstrating that eveningness is characterized by lower levels of happiness and inspiration (Önder, 2022). Beyond trait-based affect, evening-type adolescents have been observed to have lower amplitudes of positive affect than morning-types across diurnal rhythms (Carciofo, 2020; Miller et al., 2015). In physical health domains, eveningness has been linked to inflammation (Dolsen & Harvey, 2021), low physical activity (Merikanto et al., 2020), and lower levels of cardiorespiratory fitness (I. Y. Chen et al., 2024). Taken together, these findings are suggestive of risks associated with eveningness across multiple domains of development, thus meriting an exploration of protective factors for such youth.

Sleep onset consistency and adolescent development

Consistency in sleep onset has been associated with development through a general pattern that more variability in sleep onset is a unique predictor of diminished mental and physical health beyond mean sleep times (Becker et al., 2017; Sletten et al., 2023). In studies of adolescents, inconsistency in sleep is associated with externalizing problems and internalizing symptoms (Lunsford-Avery et al., 2022). Beyond behavioral and emotional factors, inconsistency in the timing of adolescent sleep onset has been linked to poorer health behaviors (Becker et al., 2017). Thus, there is empirical evidence that consistent sleep onset plays a key role in multiple domains of youth development. Furthermore, in support of potential moderation effects, both eveningness (Hasler et al., 2017) and sleep loss (a potential consequence of sleep inconsistency; Ben Simon et al., 2020) are associated with blunted reward response in the medial prefrontal cortex (mPFC), which links M – E with adolescent behavior problems, depression, and positive affect (Hasler et al., 2010, 2017).

Current study

We examined the association between eveningness and multiple domains of development and tested sleep onset consistency as a moderator of these relations. This investigation is novel in several ways. First, it examines both negative and positive indicators of development as they relate to sleep health, responsive to recent calls in the literature (Meltzer et al., 2021); we examined behavior problems, internalizing symptoms (depression, anxiety), positive affect (optimism, happiness), and a general rating of physical health. Second, to our knowledge the study is the first to examine sleep onset consistency as a protective factor for evening-type adolescents. (While we use the term evening-type for simplicity to refer to youth with a preference for more eveningness than morningness, in analyses we examined M – E on a continuum to capitalize on variability in this construct.) Such an investigation is timely given that both sleep (Rocha & Fuligni, 2023) and mental health (Panchal et al., 2023) declined globally in the wake of the COVID-19 pandemic, highlighting the need for clear evidence of which sleep behaviors youth can target to promote their mental health. Finally, weeknight and weekend mean sleep onset and weekly mean sleep minutes were covaried in analyses, allowing us to decipher unique associations of sleep onset consistency with development, over and above mean sleep onset and duration. We hypothesized that greater eveningness would be associated with suboptimal development – i.e., more behavior problems and internalizing symptoms, less optimism and happiness, and lower physical health. We further expected that sleep onset consistency would interact with M – E and attenuate each of these associations, such that consistent onset would protect against the suboptimal developmental outcomes for evening-type adolescents.

Methods

Participants

Data were drawn from the fourth wave of the Auburn University Sleep Study; for detailed recruitment and sample information, see Shimizu et al. (2021).1 Data were collected in 2017–2018 from 323 adolescents and their families residing in small towns and surrounding rural communities in the southeastern U.S.A. Adolescents did not qualify for participation if they had a diagnosed sleep disorder or learning disability based on mothers’ reports. To reduce potential confounds, the analytic sample for the present study included only participants who were physically attending high school (N = 272; Mage = 17.3 years, SD = 9.12 months; 49% female; 41% non-Hispanic Black/African-American and 59% non-HispanicWhite/European-American (referred to hereafter as Black and White). Families were socioeconomically diverse: 40% near or below the poverty line, 17% lower middle class, and 43% middle class or higher.

Procedures

The Auburn University institutional review board approved study procedures (protocol #08–296 MR 0812), and parents and adolescents provided written consent and assent. Adolescents wore actigraphs at home for seven consecutive nights during the school year and completed a diary during the same nights. Then, families visited an on-campus laboratory approximately two days (M = 1.87, SD = 5.62 days) after the sleep assessment to complete questionnaires (Qualtrics; Provo, UT). Some families completed the surveys online prior to the laboratory visit; in such cases, surveys were sent via e-mail around the same time actigraphs were mailed to their homes.

Measures

Morningness–eveningness

Adolescents completed the morningness – eveningness scale of the School Sleep Habits Survey (Wolfson & Carskadon, 1998; 10 items), which is validated among youth ages 13–19. The measure assesses the times of day that individuals prefer to wake and sleep given no other constraints, the ease with which they wake in the morning, and their alertness throughout the day. Responses could range from 1 to either 4 or 5 on various items, with higher scores indicating more morningness. Summed scores were used in analyses, with a full possible range of 10 (extreme eveningness) to 43 (extreme morningness). The morningness – eveningness scale had acceptable reliability in our sample (α = .73).

Sleep onset consistency

Data from Octagonal Basic Motionlogger actigraphs (Ambulatory Monitoring, Ardsley, NY, USA) were scored in ActionW2 using the Sadeh algorithm (Sadeh et al., 1994) and zero crossing mode to derive the number of 1-min epochs scored as sleep. A sleep diary was used to verify sleep and wake times while coding the actigraphy data; on rare instances in which actigraphy onset and wake times differed from self-reported times by more than 30 min, data were not scored in accordance with the recommendation of Meltzer et al. (2012) that actigraphy not be scored without sleep diary corroboration. Consistent with definitions in the scoring manual, two sleep parameters were obtained: sleep onset time (the first of at least 3 consecutive minutes of sleep) and sleep minutes (the number of epochs scored as sleep between sleep onset and wake time). In accordance with guidelines (Meltzer et al., 2012), sleep parameters were analyzed if participants had at least five nights of actigraph data, excluding nights when the participant reported using medication for acute illnesses. A total of 118 participants (43.38%) had data for 7 nights, 69 (25.37%) for 6 nights, and 39 (14.34%) for 5 nights. Actigraph sleep data were treated as missing for the remaining 46 participants (16.91%) though their data on other variables were retained for analyses (see Statistical Analyses). Consistency in sleep onset during weeknights was calculated using the coefficient of variation (Snedecor & Cochran, 1989), which is the intraindividual standard deviation divided by the intraindividual mean. For parity with the literature (Becker et al., 2017), consistency in sleep onset is referred to as variability in the Methods and Results.

Adolescent development

Externalizing behavior.

Mothers reported on adolescents’ externalizing behavior using the Personality Inventory for Children-2, a widely used measure validated for youth from ages 5 to 19 (Lachar & Gruber, 2001). The composite externalizing scale consists of 24 items (Yes/No) assessing behaviors of aggression, delinquency, impulsivity, noncompliance, and disruption (α =.89). T-scores, normed for age and sex, were used for analysis (α =.89 in our sample). Few participants (9.72%) had clinically significant scores.

Internalizing symptoms.

Adolescents reported on their internalizing symptoms using the well-established Children’s Depression Inventory (CDI), which is validated from ages 7 to 17 (Kovacs, 1992) and the Revised Children’s Manifest Anxiety Scale (RCMAS), which is validated for youth ages 6–19 (Reynolds & Richmond, 2008). Adolescent-report measures were selected for internalizing symptoms given that youth are more accurate reporters of their own depression and anxiety than their parents (Dowdy & Kim, 2012). The CDI includes 27 items, indicating the degree to which individuals experienced symptoms of depression over the previous two weeks across statements such as, “Things will work out for me O.K.” (0), “I am not sure if things will work out for me” (1), and “Nothing will ever work out for me” (2). Two questions pertaining to sleep disturbances were excluded to avoid conflation between independent and dependent variables; the modified measure was reliable in our sample (α = .88). Summed scores could range from 0 to 50. Six percent of participants met a clinical cutoff for depression screening (scores ≥20; Kovacs, 1992).

The total anxiety scale of the RCMAS includes 40 items assessing three dimensions of anxiety: physiology, worry, and social anxiety. Adolescents reported whether they agreed (1) or disagreed (0) with statements such as, “I worry about what is going to happen,” and “I fear other kids will laugh at me in class.” Five items surveying sleep problems were excluded, producing a modified scale with high reliability in our sample (α = .93) and summed scores could range from 0 to 35. Three percent of participants had anxiety symptoms in the clinical range (≥2 SD from the mean; Reynolds & Richmond, 2008). The CDI and RCMAS were highly correlated (r = .76) and were standardized and summed to create a composite of internalizing symptoms.

Positive affect.

Adolescents reported on their positive affect using the Subjective Happiness Scale (SHS), which is validated for use among students aged 14 to 28 (Lyubomirsky & Lepper, 1999), and the Youth Life Orientation Test (YLOT; Ey et al., 2005), which has been validated among youth ages 13 to 16 (Mendes de Oliveira et al., 2022) and used among youth through age 18 (Fairbank et al., 2024). The SHS and the YLOT are well-established measures of youths’ psychological wellbeing. The SHS consists of 4 items (α = .76 in our sample) assessing overall subjective happiness, with responses from 1 (Not happy/Not at all) to 7 (Very happy/Great), averaged for a possible range of 1 to 7. Two items ask the degree to which respondents consider themselves a happy individual in general and relative to peers, and the other two items offer brief descriptions of happy and unhappy individuals and ask respondents the extent to which each characterization describes them.

The YLOT consists of 12 items that include 6 on optimism (e.g., “I usually expect to have a good day”) and 6 on pessimism (e.g., “Things usually go wrong for me”), each with responses from 0 (Not true for me) to 3 (True for me), summed for a possible range of 0 to 36. The YLOT total optimism score was calculated by summing the optimism and reverse-scored pessimism items (α = .86 in our sample), with higher scores indicating higher optimism. The SHS and the YLOT total optimism scores were highly correlated (r = .64) and were standardized and summed to create a composite score of positive affect.

Physical health.

Adolescents reported on their general physical health using a single item rating from 1 (Poor) and 5 (Excellent) from the Child Health and Illness Profile – Adolescent Edition (Starfield et al., 1993), which is validated for use in ages 11 to 17. Such single-item ratings of health are shown to have high correspondence to multiple-item ratings (Verster et al., 2021) and to objective health indicators (Schnittker et al., 2014).

Covariates

Mother-reported youth sex and race, and family socioeconomic status (SES) indexed by family income-to-needs ratio (income divided by the federal poverty threshold for household size; U.S. Department of Commerce, 2017), were covaried in analyses. To examine the unique effects of variability in weeknight sleep onset on adolescent development beyond average times and beyond sleep duration, the mean times of weekday and weekend sleep onset, as well as mean sleep minutes for all nights, were included as control variables.

Statistical analyses

A series of four multiple regression models were fit. M – E, consistency in sleep onset, and covariates (adolescent sex and race/ethnicity, family SES, weeknight mean sleep onset, weekend mean sleep onset, and weekly mean sleep minutes) were entered simultaneously and mean-centered to facilitate interpretation of intercepts. Further, we covaried exogenous variables that were significantly related in bivariate correlations. Interactions were plotted at high (+1 SD) and low (−1 SD) levels of M – E and weeknight variability in sleep onset using an online tool (Preacher et al., 2006). Simple slopes, chi-square difference testing, and regions of significance were used to interpret interaction plots.

All models were fit in Amos 25 (Amos Development Corporation, Wexford, PA, USA) using full information maximum likelihood estimation (FIML) to handle missing data (Enders, 2001). Missingness for study variables ranged from 2.21% to 22.05%, which is within the acceptable range for FIML (McNeish, 2017).2 For variables with skewness values greater than 2.00, Mahalanobis distance was used to guide the trimming of influential outliers (Kline, 2011); a total of six observations were trimmed to the next-highest observed value. All models had acceptable fit by meeting at least two of the following three criteria: χ2/df < 2, comparative fit index >0.95, and root mean square error of approximation < 0.06 (Hu & Bentler, 1999).

Results

Preliminary analyses

Adolescents had an average sleep onset time of 11:13 pm on weeknights and 12:19 am on weekends, roughly on par with similarly aged youth globally (Gariépy et al., 2020). Youth slept an average of 6 hr, 37 min on weeknights and 7 hr, 6 min on weekends, excluding minutes of wakefulness after sleep onset; this is shorter than most other youth (Gariépy et al., 2020) and below recommendations (Hirshkowitz et al., 2015; although, notably, recommendations are based on self-reported sleep duration inclusive of wakefulness between sleep onset and offset). Table 1 shows additional descriptive statistics of sleep. M – E and variability in sleep onset did not vary by sex, race, or SES. For outcome variables, externalizing behavior did not vary by sex, but females reported more internalizing symptoms (females: M = 0.40, SD = 2.08; males: M = −0.39, SD = 1.56), t(229.40) = 3.41, p = .001; less positive affect (females: M = −0.25, SD = 1.82; males: M = 0.25, SD = 1.78), t(212.00) = −2.07, p < .05; and poorer physical health (females: M = 3.66, SD = 0.97; males: M = 4.11, SD = 0.78), t(194.15) = −3.72, p < .001 than males. Only two indicators of adolescent development varied by race or SES: Black youth (M = −0.30, SD = 1.71) reported fewer internalizing symptoms than White peers (M = 0.19, SD = 1.94), t(258) = 2.09, p < .05, and lower-SES adolescents displayed more externalizing problems than higher-SES youth, r = −0.19, p < .01. Table 2 displays correlations among main study variables.

Table 1.

Descriptive statistics of continuous variables.

n M SD Min. Max.
Main study variables
 Morningness – eveningness 261 27.20 4.76 12.00 38.00
 Weeknight variability in sleep onset 250 3.46 2.04 0.34 10.69
 Externalizing behavior 247 46.74 8.54 38.00 81.00
 Internalizing symptoms 260 −0.01 1.86 −2.39 7.06
 Positive affect 214 0.01 1.82 −5.41 3.51
 Physical health 212 3.89 0.91 1.00 5.00
Covariates
 Mean weeknight sleep onset 250 11:13pm 63.62 min 7:28pm 2:31am
 Mean weekend sleep onset 237 12:19am 97.62 min 7:54pm 8:01am
 Mean sleep minutes (all nights) 226 407.18 59.30 214.00 614.43
 Income-to-needs ratio 269 2.88 1.74 0.15 7.84
Additional sleep parameters
 Mean weeknight sleep offset 247 6:18am 50.74 min 4:04am 9:47am
 Mean weekend sleep offset 232 8:01am 85.75 min 4:11am 12:24pm
 Sleep efficiency 226 93.34% 6.29% 62.91% 100.00%
 Wake after sleep onset 226 29.02 27.61 0.00 172.00
 Long wake episodes (>5 min) 226 1.83 1.48 0.00 9.40
Open in a new tab

Variability in sleep onset represented as the coefficient of variation (intra-individual SD/M) multiplied by 100 to increase variance for model convergence. Mean sleep minutes is equivalent to total sleep time. 407.18 min = 6 hours, 47 min. “Additional sleep parameters” were measured across all nights except where indicted; these parameters were not utilized in models but are presented for informative and comparative purposes. Sleep efficiency calculated as sleep minutes [total sleep time] as a percentage of the sleep period (elsewhere referred to as sleep percent; Ancoli-Israel et al., 2015). Sleep quality means (efficiency, wake after sleep onset, and long wake episodes) are within range for youth of this age (Ohayon et al., 2017).

Table 2.

Correlations among main study variables.

1 2 3 4 5 6
1. Morningness – eveningness —
2. Variability in sleep onset −0.04 —
3. Externalizing behavior −0.14* 0.20** —
4. Internalizing symptoms −0.34*** 0.19** 0.25*** —
5. Positive affect 0.29*** −0.13 −0.22** −0.73*** —
6. Physical health 0.13 −0.15* −0.15* −0.36*** 0.39*** —
Open in a new tab
*

p < .05.

**

p < .01.

***

p ≤ .001.

Associations of morningness–eveningness and variability in sleep onset with development

Eveningness (i.e., lower score on the Morningness – Eveningness Scale) was associated with more externalizing problems (B = −0.27, p < .05), greater internalizing symptoms (B = −0.14, p < .001), less positive affect (B = 0.12, p < .001), and poorer physical health (B = 0.03, p < .05; Table 3). More variability in sleep onset was associated with more externalizing problems, greater internalizing symptoms, less positive affect, and poorer physical health.

Table 3.

Associations of morningness–eveningness and weeknight variability in sleep onset with adolescent development.

Externalizing Behavior
 Internalizing Symptoms
 Positive Affect
 Physical Health
B SE B SE B SE B SE
Intercept 46.74*** 0.51 0.003 0.10 −0.01 0.11 3.88*** 0.06
Morningness – eveningness −0.27* 0.11 −0.14*** 0.02 0.12*** 0.03 0.03* 0.01
Weeknight variability in sleep onset 0.84*** 0.25 0.16** 0.05 −0.11* 0.06 −0.07* 0.03
M – E x Weeknight Variability in Sleep Onset −0.13* 0.06 −0.02~ 0.01 0.02~ 0.01 0.01~ 0.01
Mean weeknight sleep onset 0.004 0.01 −0.002 0.003 0.00 0.003 −0.001 0.001
Mean weekend sleep onset −0.01 0.01 0.00 0.001 0.001 0.002 0.001 0.001
Mean sleep minutes (all nights) −0.01 0.01 −0.001 0.002 −0.003 0.003 0.00 0.001
Sex −1.35 1.03 −0.70*** 0.21 0.32 0.23 0.41*** 0.12
Race −1.70 1.17 −0.75*** 0.24 0.54* 0.26 0.21 0.13
Income-to-needs ratio −0.97** 0.33 −0.12 0.07 0.15* 0.07 0.01 0.04
R 2 0.14 0.22 0.18 0.14
Open in a new tab

N = 272. M – E = Morningness – eveningness. Models were fit separately for each outcome. Mean sleep minutes is equivalent to total sleep time.

~

p ≤ .10.

*

p < .05.

**

p < .01.

***

p ≤ .001.

Moderation by variability in sleep onset

Variability in sleep onset moderated associations between morningness – eveningness and externalizing behavior, internalizing symptoms, positive affect, and physical health at the level of a statistical trend or higher (Table 3).

Externalizing behavior

Simple slope analyses indicated a significant association between M – E and externalizing behavior only for youth with more variability in sleep onset (Figure 1a). Among evening-type youth, those with greater variability in sleep onset showed externalizing behavior above the mean, Δχ2 (1) = 64.51, p < .001, while those with less variability in sleep onset had externalizing behavior below the mean, Δχ2 (1) = 11.83, p < .001. Regions of significance showed that the association between eveningness and externalizing was significant at a coefficient of variation in sleep onset of 0.03 or higher; 46.80% of the sample fell in this range. The interaction of M – E x Variability in Sleep Onset accounted for 2.6% of the variance in externalizing.

Figure 1.

Figure 1.

Open in a new tab

Relations between morningness–eveningness and adolescent development moderated by weeknight variability in sleep onset. Prototypical slopes of the relationship between morningness – eveningness and adolescent development plotted at 1 SD above and below the mean for morningness – eveningness and weeknight variability in sleep onset. Coefficients for simple slopes are indicated. Panel A: externalizing behavior. Panel B: internalizing symptoms. Panel C: positive affect. Panel D: physical health.

*p < .05. **p < .01. ***p < .001.

Internalizing symptoms

An examination of simple slopes revealed that the relationship between M – E and internalizing symptoms was significant for adolescents with both higher and lower variability in sleep onset, with a steeper slope for the former (Figure 1b). Evening-type youth with more variability in sleep onset had a predicted mean of internalizing symptoms that was greater than the mean for the full sample, Δχ2 (1) = 106.55, p < .001, whereas evening-type youth with less variability in sleep onset did not differ significantly from the mean, Δχ2 (1) = 1.83, p = .18. Regions of significance showed that the relationship between M – E and internalizing was significant for youth with variability in sleep onset of 0.01 or above (98.80% of the sample). The interaction term accounted for 1.3% of the variance in internalizing.

Positive affect

Analyses of simple slopes showed that the association of M – E with positive affect was significant at both higher and lower levels of variability in sleep onset; however, the slope was steeper for youth with higher variability (Figure 1c). Among evening-type youth, those with more variability in sleep onset had positive affect below the mean, Δχ2 (1) = 69.40, p < .001, while those with less variability in sleep onset had positive affect not significantly different from the mean, Δχ2 (1) = 1.61, p = .20. The relationship between M – E and positive affect was significant at a variability in sleep onset of 0.01 or above (89.60% of sample). Two percent of the variance in positive affected was accounted for by the interaction term.

Physical health

Simple slope analyses revealed a relationship between M – E and physical health only for youth with more variability in sleep onset (Figure 1d). Among evening-type youth, those with more variability in sleep onset had poorer physical health, significantly below the mean, Δχ2 (1) = 41.65, p < .001, while those with less variability had physical health not significantly different from the mean, Δχ2 (1) = 3.65, p = .06. The association between eveningness and poorer physical health was significant at a variability in sleep onset of 0.03 or above (47.60% of the sample). The interaction term accounted for 2.00% of the variance in physical health.

Discussion

Sleep onset consistency is associated with adolescent development (Bei et al., 2016), yet this association has not been explored in the context of circadian preference. Contributing to the literature in novel ways, we examined whether consistency in weeknight sleep onset is protective against negative outcomes and promotes positive outcomes among night owls among a community sample of non-sleep-disordered adolescents. Supportive of hypotheses, consistency in actigraphy-assessed sleep onset moderated relations between M – E and development.

A protective pattern for night owls emerged across outcomes: evening-type youth who were more consistent in weeknight sleep onset times had lower predicted behavior and emotion problems and higher predicted positive affect and physical health, on par with their morning-type peers and sample means. Though three interactions were significant only at the level of p ≤ .10, interaction plots, together with tests of simple slopes and regions of significance, affirmed that the association between M – E and all four indicators of development varied by sleep onset consistency.

Examination of variance in sleep onset consistency showed that 1 SD below the mean – the level of sleep onset consistency estimated to be protective in the prototypical plots in Figure 1 – equates to falling asleep within approximately 30 min of the same time each night (15 min before or after). This is consistent with other recommendations for youth (Allen et al., 2016) and is in line with the National Sleep Foundation’s consensus statement on the importance of sleep regularity (Sletten et al., 2023), although such a statement has not yet been made for pediatric sleep.

It is well documented that high school students stay awake and sleep later on weekends than weeknights (Henderson et al., 2019; Nicholson et al., 2023). Our study isolated variability in weeknight sleep onset while covarying both weeknight and weekend mean onset. Our findings suggest that even if night owls fall asleep later than morning-types during the week – and later still on the weekend – their weeknight sleep onset consistency was still protective of mental and physical health. Such an approach – which takes into account normative youth behaviors around sleep onset – is encouraged in the pediatric sleep health literature (Meltzer et al., 2021).

It is notable that youth of this age commonly experience anxiety, including social anxiety (Parodi et al., 2022). Sleep consistency and anxiety are reciprocally related to one another (Kelly et al., 2022). Our findings suggest that such relations could be stronger for evening-type youth compared to morning-type youth, who, in our sample, had lower internalizing symptoms regardless of sleep onset consistency.

As noted above, the present study’s independent and dependent variables are associated with activity in the mPFC, suggesting a potential, though tentative, mechanism for the relationships we observed. Sleep onset consistency primes optimal reward response (Zhang et al., 2020), which may counteract lower regulation among evening-type adolescents (Hasler et al., 2017). This may help to explain why evening-type youth with more consistent sleep onset in our sample had fewer externalizing behaviors and less internalizing symptoms than evening-type youth with less consistent sleep onset.

Strengths of the study include its broad distribution of M – E and sleep onset consistency, as well its use of actigraphy. Generalizability of findings is limited by small effect sizes and restricted to the sociodemographic characteristics of the sample, which was representative of the recruitment area regarding race/ethnicity and family income. Findings may vary in urban communities, where youth have later weeknight sleep timing compared to adolescents in more rural areas (Kim et al., 2020). The cross-sectional nature of the data prohibits causal inference. Our measure of physical health, a single-item self-report, was limited by a high ceiling, which is common among healthy youth (Waters et al., 2001); nonetheless, physical health varied by M – E in our sample, and this association was moderated by sleep onset consistency, reflective of meaningful distribution of health in the sample. Future work may model the effects of changes in sleep onset consistency on relations between M – E and adolescent development both across weeks, months, and years, and within days (Bettencourt et al., 2022). Our study offers an important message to parents, youth, and health professionals who safeguard the behavioral, emotional, and physical health of adolescents: consistency in sleep onset on weeknights is promotive of wellbeing for adolescent night owls.

Acknowledgments

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors wish to thank laboratory staff and the families who participated.

Funding

The work was supported by the National Heart, Lung, and Blood Institute [R01-HL136752, R01-HL093246].

Footnotes

Disclosure statement

No potential conflict of interest was reported by the author(s).

1

We selected the fourth wave of this study because we wanted to examine findings that could be used by adolescents to improve or protect their own health. In the first three waves of the study, children were 9, 10, and 11 years old on average, respectively; at these ages, sleep schedules are frequently set by parents (Pyper et al., 2017); by late adolescence, teens have more autonomy over bed times and thus could utilize these findings (Peltz et al., 2020). Conversely, by the fifth wave of data, half of the sample had completed secondary school, so their sleep schedules were no longer comparable to the participants still adhering to secondary school schedules.

2

To ensure confidence in results, we conducted sensitivity analyses excluding participants with missing data. Results did not differ substantively from the findings presented here.

Data availability statement

Data are not available yet for sharing with others. Per National Institutes of Health data sharing guidelines, they will be available to other scholars at a later date after the completion of this longitudinal study.

References

  1. Adan A, Archer SN, Hidalgo MP, DiMilia L, Natale V, & Randler C. (2012). Circadian typology: A comprehensive review. Chronobiology International, 29(9), 1153–1175. 10.3109/07420528.2012.719971 [DOI] [PubMed] [Google Scholar]
  2. Allen SL, Howlett MD, Coulombe JA, & Corkum PV (2016). ABCs of SLEEPING: A review of the evidence behind pediatric sleep practice recommendations. Sleep Medicine Reviews, 29, 1–14. 10.1016/j.smrv.2015.08.006 [DOI] [PubMed] [Google Scholar]
  3. Ancoli-Israel S, Martin JL, Blackwell T, Buenaver L, Liu L, Meltzer LJ, Sadeh A, Spira AP, & Taylor DJ (2015). The SBSM guide to actigraphy monitoring: Clinical and research applications. Behavioral Sleep Medicine, 13 (sup1), S4–S38. 10.1080/15402002.2015.1046356 [DOI] [PubMed] [Google Scholar]
  4. Azad Marzabadi E, & Amiri S. (2017). Morningness–eveningness and emotion dysregulation incremental validity in predicting social anxiety dimensions. International Journal of General Medicine, 10, 275–279. 10.2147/IJGM.S144376 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Becker SP, Sidol CA, Van Dyk TR, Epstein JN, & Beebe DW (2017). Intraindividual variability of sleep/wake patterns in relation to child and adolescent functioning: A systematic review. Sleep Medicine Reviews, 34, 94–121. 10.1016/j.smrv.2016.07.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bei B, Wiley JF, Trinder J, & Manber R. (2016). Beyond the mean: A systematic review on the correlates of daily intraindividual variability of sleep/wake patterns. Sleep Medicine Reviews, 28, 108–124. 10.1016/j.smrv.2015.06.003 [DOI] [PubMed] [Google Scholar]
  7. Ben Simon E, Vallat R, Barnes CM, & Walker MP (2020). Sleep loss and the socio-emotional brain. Trends in Cognitive Sciences, 24(6), 435–450. 10.1016/j.tics.2020.02.003 [DOI] [PubMed] [Google Scholar]
  8. Bettencourt C, Tomé B, Pires L, Leitão JA, & Gomes AA (2022). Emotional states in adolescents: Time of day X chronotype effects while controlling for psychopathological symptoms and sleep variables. Biological Rhythm Research, 53(3), 478–499. 10.1080/09291016.2020.1783489 [DOI] [Google Scholar]
  9. Carciofo R. (2020). Morningness–eveningness and affect: The mediating roles of sleep quality and metacognitive beliefs. Sleep and Biological Rhythms, 18(1), 17–26. 10.1007/s41105-019-00238-9 [DOI] [Google Scholar]
  10. Chen IY, Radom-Aizik S, Stehli A, Palmer JR, Lui KK, Dave A, Chappel-Farley MG, Vinces KG, Gealer D, Lim A, Mander BA, Benca RM, & Neikrug AB (2024). Cardiorespiratory fitness and circadian rhythms in adolescents: A pilot study. Journal of Applied Physiology, 136(2), 372–384. 10.1152/japplphysiol.00291.2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chen S-J, Zhang J-H, Li SX, Tsang CC, Chan KCC, Au CT, Li AM, Kong APS, Wing YK, & Chan NY (2021). The trajectories and associations of eveningness and insomnia with daytime sleepiness, depression and suicidal ideation in adolescents: A 3-year longitudinal study. Journal of Affective Disorders, 294, 533–542. 10.1016/j.jad.2021.07.033 [DOI] [PubMed] [Google Scholar]
  12. Cheung FTW, Li X, Hui TK, Chan NY, Chan JW, Wing YK, & Li SX (2023). Circadian preference and mental health outcomes in youth: A systematic review and meta-analysis. Sleep Medicine Reviews, 72, 101851. 10.1016/j.smrv.2023.101851 [DOI] [PubMed] [Google Scholar]
  13. Cooper R, DiBiase MA, Bei B, Allen NB, Schwartz O, Whittle S, & Cropley V. (2023). Development of morning–eveningness in adolescence: Implications for brain development and psychopathology. Journal of Child Psychology and Psychiatry, 64(3), 449–460. 10.1111/jcpp.13718 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dolsen EA, & Harvey AG (2021). IL-6, sTNF-R2, and CRP in the context of sleep, circadian preference, and health in adolescents with eveningness chronotype: Cross-sectional and longitudinal treatment effects. Psychoneuroendocrinology, 129, 105241. 10.1016/j.psyneuen.2021.105241 [DOI] [PubMed] [Google Scholar]
  15. Dowdy E, & Kim E. (2012). Choosing informants when conducting a universal screening for behavioral and emotional risk. School Psychology Forum: Research in Practice, 6(4), 1–10. [Google Scholar]
  16. Enders CK (2001). The performance of the full information maximum likelihood estimator in multiple regression models with missing data. Educational and Psychological Measurement, 61(5), 713–740. 10.1177/0013164401615001 [DOI] [Google Scholar]
  17. Ey S, Hadley W, Allen DN, Palmer S, Klosky J, Deptula D, Thomas J, & Cohen R. (2005). A new measure of children’s optimism and pessimism: The youth life orientation test. Journal of Child Psychology and Psychiatry, 46(5), 548–558. 10.1111/j.1469-7610.2004.00372.x [DOI] [PubMed] [Google Scholar]
  18. Fairbank EJ, Borenstein-Laurie J, Alberts NM, & Wrosch C. (2024). Optimism, pessimism, and physical health among youth: A scoping review. Journal of Pediatric Psychology, jsae045. 10.1093/jpepsy/jsae045 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gariépy G, Danna S, Gobiņa I, Rasmussen M, de Matos MG, Tynjälä J, Janssen I, Kalman M, Villeruša A, Husarova D, Brooks F, Elgar FJ, Klavina-Makrecka S, Šmigelskas K, Gaspar T, & Schnohr C. (2020). How are adolescents sleeping? Adolescent sleep patterns and sociodemographic differences in 24 European and North American countries. Journal of Adolescent Health, 66(6), S81–S88. 10.1016/j.jadohealth.2020.03.013 [DOI] [PubMed] [Google Scholar]
  20. Hasler BP, Allen JJ, Sbarra DA, Bootzin RR, & Bernert RA (2010). Morningness–eveningness and depression: Preliminary evidence for the role of the behavioral activation system and positive affect. Psychiatry Research, 176 (2–3), 166–173. 10.1016/j.psychres.2009.06.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hasler BP, Casement MD, Sitnick SL, Shaw DS, & Forbes EE (2017). Eveningness among late adolescent males predicts neural reactivity to reward and alcohol dependence 2 years later. Behavioural Brain Research, 327, 112–120. 10.1016/j.bbr.2017.02.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Henderson SE, Brady EM, & Robertson N. (2019). Associations between social jetlag and mental health in young people: A systematic review. Chronobiology International, 36(10), 1316–1333. 10.1080/07420528.2019.1636813 [DOI] [PubMed] [Google Scholar]
  23. Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, Hazen N, Herman J, Adams Hillard PJ, Katz ES, Kheirandish-Gozal L, Neubauer DN, O’Donnell AE, Ohayon M, Peever J, Rawding R, Sachdeva RC, Setters B, Vitiello MV, & Ware JC (2015). National Sleep Foundation’s updated sleep duration recommendations: Final report. Sleep Health, 1(4), 233–243. 10.1016/j.sleh.2015.10.004 [DOI] [PubMed] [Google Scholar]
  24. Hu LT, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  25. Jankowski KS, & Linke M. (2020). Angry night birds: Emotionality, activity and sociability temperament in adolescent chronotypes. Chronobiology International, 37(5), 652–659. 10.1080/07420528.2020.1754844 [DOI] [PubMed] [Google Scholar]
  26. Kelly RJ, Zeringue MM, & El-Sheikh M. (2022). Adolescents’ sleep and adjustment: Reciprocal effects. Child Development, 93(2), 540–555. 10.1111/cdev.13703 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kim J, Noh J-W, Kim A, & Kwon YD (2020). Demographic and socioeconomic influences on sleep patterns among adolescent students. International Journal of Environmental Research and Public Health, 17(12), 4378. 10.3390/ijerph17124378 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kline RB (2011). Principles and practice of structural equation modeling (3rd ed.) Guilford Press. [Google Scholar]
  29. Kovacs M. (1992). Children’s Depression Inventory. Multi-Health Systems. [Google Scholar]
  30. Lachar D, & Gruber CP (2001). Personality Inventory for Children: Second Edition (PIC-2). Western Psychological Services. [Google Scholar]
  31. Lunsford-Avery JR, Wang KW, Kollins SH, Chung RJ, Keller C, & Engelhard MM (2022). Regularity and timing of sleep patterns and behavioral health among adolescents. Journal of Developmental & Behavioral Pediatrics, 43(4), 188–196. 10.1097/DBP.0000000000001013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lyubomirsky S, & Lepper HS (1999). A measure of subjective happiness: Preliminary reliability and construct validation. Social Indicators Research, 46(2), 137–155. 10.1023/a:1006824100041 [DOI] [Google Scholar]
  33. McNeish D. (2017). Missing data methods for arbitrary missingness with small samples. Journal of Applied Statistics, 44 (1), 24–39. 10.1080/02664763.2016.1158246 [DOI] [Google Scholar]
  34. Meltzer LJ, Montgomery-Downs HE, Insana SP, & Walsh CM (2012). Use of actigraphy for assessment in pediatric sleep research. Sleep Medicine Reviews, 16(5), 463–475. 10.1016/j.smrv.2011.10.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Meltzer LJ, Williamson AA, & Mindell JA (2021). Pediatric sleep health: It matters, and so does how we define it. Sleep Medicine Reviews, 57, 101425. 10.1016/j.smrv.2021.101425 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Mendes de Oliveira C, Zanon C, de Moraes Bandeira C, Heath PJ, & Hofheinz Giacomoni C. (2022). Evaluating optimism in children and adolescents: Adaptation, factor structure, convergent validity, and invariance of the Brazilian version of the Youth Life Orientation Test (YLOT). Psychological Assessment, 34(1), e1. 10.1037/pas0001090 [DOI] [PubMed] [Google Scholar]
  37. Merikanto I, Kuula L, Lahti J, Räikkönen K, & Pesonen A-K (2020). Eveningness associates with lower physical activity from pre-to late adolescence. Sleep Medicine, 74, 189–198. 10.1016/j.sleep.2020.07.021 [DOI] [PubMed] [Google Scholar]
  38. Miller MA, Rothenberger SD, Hasler BP, Donofry SD, Wong PM, Manuck SB, Kamarck TW, & Roecklein KA (2015). Chronotype predicts positive affect rhythms measured by ecological momentary assessment. Chronobiology International, 32(3), 376–384. 10.3109/07420528.2014.983602 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Nicholson L, Bohnert AM, & Crowley SJ (2023). A developmental perspective on sleep consistency: Preschool age through emerging adulthood. Behavioral Sleep Medicine, 21(1), 97–116. 10.1080/15402002.2021.2024192 [DOI] [PubMed] [Google Scholar]
  40. Ohayon M, Wickwire EM, Hirshkowitz M, Albert SM, Avidan A, Daly FJ, Dauvilliers Y, Ferri R, Fung C, Gozal D, Hazen N, Krystal A, Lichstein K, Mallampalli M, Plazzi G, Rawding R, Scheer FA, Somers V, & Vitiello MV (2017). National Sleep Foundation’s sleep quality recommendations: First report. Sleep Health, 3(1), 6–19. 10.1016/j.sleh.2016.11.006 [DOI] [PubMed] [Google Scholar]
  41. Önder İ (2022). Association of happiness with morningness-eveningness preference, sleep-related variables and academic performance in university students. Biological Rhythm Research, 53(6), 950–965. 10.1080/09291016.2020.1848266 [DOI] [Google Scholar]
  42. Panchal U, Salazar de Pablo G, Franco M, Moreno C, Parellada M, Arango C, & Fusar-Poli P. (2023). The impact of COVID-19 lockdown on child and adolescent mental health: Systematic review. European Child & Adolescent Psychiatry, 32(7), 1151–1177. 10.1007/s00787-021-01856-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Panjeh S, Pompeia S, Archer SN, Pedrazzoli M, von Schantz M, & Cogo-Moreira H. (2021). What are we measuring with the morningness–eveningness questionnaire? Exploratory factor analysis across four samples from two countries. Chronobiology International, 38(2), 234–247. 10.1080/07420528.2020.1815758 [DOI] [PubMed] [Google Scholar]
  44. Parodi KB, Holt MK, Green JG, Porche MV, Koenig B, & Xuan Z. (2022). Time trends and disparities in anxiety among adolescents, 2012–2018. Social Psychiatry & Psychiatric Epidemiology, 57(1), 127–137. 10.1007/s00127-021-02122-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Peltz JS, Rogge RD, & Connolly H. (2020). Parents still matter: The influence of parental enforcement of bedtime on adolescents’ depressive symptoms. Sleep, 43(5), zsz287. 10.1093/sleep/zsz287 [DOI] [PubMed] [Google Scholar]
  46. Preacher KJ, Curran PJ, & Bauer DJ (2006). Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics, 31(4), 437–448. 10.3102/10769986031004437 [DOI] [Google Scholar]
  47. Pyper E, Harrington D, & Manson H. (2017). Do parents’ support behaviours predict whether or not their children get sufficient sleep? A cross-sectional study. BMC Public Health, 17(1), 432. 10.1186/s12889-017-4334-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Reynolds CR, & Richmond BO (2008). Revised Children’s Manifest Anxiety Scale (RCMAS-2) (2nd ed.) Western Psychological Services. [Google Scholar]
  49. Rocha S, & Fuligni A. (2023). The impact of the COVID-19 pandemic on adolescent sleep behavior. Current Opinion in Psychology, 52, 101648. 10.1016/j.copsyc.2023.101648 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Sadeh A, Sharkey KM, & Carskadon MA (1994). Activity-based sleep-wake identification: An empirical test of methodological issues. Sleep, 17(3), 201–207. 10.1093/sleep/17.3.201 [DOI] [PubMed] [Google Scholar]
  51. Schnittker J, Bacak V, & Noymer A. (2014). The increasing predictive validity of self-rated health. PLOS ONE, 9(1), e84933. 10.1371/journal.pone.0084933 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Shimizu M, Zeringue MM, Erath SA, Hinnant JB, & El-Sheikh M. (2021). Trajectories of sleep problems in childhood: Associations with mental health in adolescence. Sleep, 44(3), zsaa190. 10.1093/sleep/zsaa190 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Sletten TL, Weaver MD, Foster RG, Gozal D, Klerman EB, Rajaratnam SM, Roenneberg T, Takahashi JS, Turek FW, Vitiello MV, Young MW, & Czeisler CA (2023). The importance of sleep regularity: A consensus statement of the national sleep foundation sleep timing and variability panel. Sleep Health, 9(6), 801–820. 10.1016/j.sleh.2023.07.016 [DOI] [PubMed] [Google Scholar]
  54. Snedecor GW, & Cochran WG (1989). Statistical methods (8th ed.). Iowa State University Press. [Google Scholar]
  55. Starfield B, Ensminger M, Riley A, McGauhey P, Skinner A, Kim S, Bergner M, Ryan S, & Green B. (1993). Adolescent health status measurement: Development of the Child Health and Illness Profile. Pediatrics, 91(2), 430–435. 10.1542/peds.91.2.430 [DOI] [PubMed] [Google Scholar]
  56. Tosuntaş ŞB, Balta S, Emirtekin E, Kircaburun K, & Griffiths MD (2020). Adolescents’ eveningness chronotype and cyberbullying perpetration: The mediating role of depression-related aggression and anxiety-related aggression. Biological Rhythm Research, 51(1), 40–50. 10.1080/09291016.2018.1513132 [DOI] [Google Scholar]
  57. Verster JC, Sandalova E, Garssen J, & Bruce G. (2021). The use of single-item ratings versus traditional multiple-item questionnaires to assess mood and health. European Journal of Investigation in Health, Psychology and Education, 11(1), 183–198. 10.3390/ejihpe11010015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Waters EB, Salmon LA, Wake M, Wright M, & Hesketh KD (2001). The health and well-being of adolescents: A school-based population study of the self-report child health questionnaire. Journal of Adolescent Health, 29(2), 140–149. 10.1016/S1054-139X(01)00211-7 [DOI] [PubMed] [Google Scholar]
  59. Wolfson AR, & Carskadon MA (1998). Sleep schedules and daytime functioning in adolescents. Child Development, 69(4), 875–887. 10.1111/j.1467-8624.1998.tb06149.x [DOI] [PubMed] [Google Scholar]
  60. Zhang R, Tomasi D, Shokri-Kojori E, Wiers CE, Wang G-J, & Volkow ND (2020). Sleep inconsistency between weekends and weekdays is associated with changes in brain function during task and rest. Sleep, 43(10), zsaa076. 10.1093/sleep/zsaa076 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Data are not available yet for sharing with others. Per National Institutes of Health data sharing guidelines, they will be available to other scholars at a later date after the completion of this longitudinal study.

RESOURCES