Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Oct 1.
Published in final edited form as: J Sleep Res. 2023 Mar 20;32(5):e13886. doi: 10.1111/jsr.13886

Morning perception of sleep, stress, and mood, and its relationship with overnight physiological sleep: Findings from the National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA) study

Benedetta Albinni 1,2, Fiona C Baker 1,3, Harold Javitz 1, Brant P Hasler 4, Peter L Franzen 4, Duncan B Clark 4, Massimiliano de Zambotti 1,*
PMCID: PMC10509318  NIHMSID: NIHMS1899088  PMID: 36941027

Abstract

This cross-sectional study investigates objective-subjective sleep discrepancies and the physiological basis for morning perceptions of sleep, mood, and readiness, in adolescents. Data collected during a single in-lab polysomnographic (PSG) assessment from 137 healthy adolescents (61 girls; age range: 12-21 years) in the US National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA) study were analyzed. Upon awakening, participants completed questionnaires assessing sleep quality, mood, and readiness. We evaluated the relationship between overnight PSG, electroencephalographic (EEG), sleep-autonomic nervous system (ANS) functioning measures, and next morning self-reported indices. Results showed that older adolescents reported more awakenings, yet they perceived their sleep to be deeper and less restless than younger adolescents. Prediction models including sleep physiology measures (PSG, EEG, and ANS) explained between 3% and 29% of morning sleep perception, mood, and readiness indices. The subjective experience of sleep is a complex phenomenon with multiple components. Distinct physiological sleep processes contribute to the morning perception of sleep and related measures of mood and readiness. More than 70% of the variance (based on a single observation per person) in the perception of sleep, mood, and morning readiness is not explained by overnight sleep-related physiological measures, suggesting that other factors are important for the subjective sleep experience.

Keywords: Sleep quality, adolescence, sex differences, polysomnography, heart rate variability

1. Introduction

Sleep quality is an important component of health. The most clinically representative example in which perceived sleep quality is critical is in the diagnosis and manifestation of insomnia disorder. Objective sleep alterations are not necessary for a diagnosis and empirical data indicate that insomnia sufferers frequently complain of poor sleep even when objective sleep indices appear relatively normal (Harvey & Tang, 2012). Adult insomnia sufferers (Krystal & Edinger, 2008) tend to overestimate objective polysomnographic (PSG) measures of sleep onset latency (SOL) and the amount of time spent awake after sleep onset (WASO), while they underestimate the amount of total sleep time (TST). These objective-subjective sleep discrepancies are among the factors considered in both diagnostic (Edinger & Krystal, 2003) and treatment aspects of insomnia disorder, as well as in its pathophysiology.

Several empirical studies have evaluated to what extent overnight physiological sleep processes could explain morning self-reported sleep outcomes. Overnight PSG measures of sleep quality and continuity (e.g., number of awakenings, total sleep time) have been found to be the best predictors of morning self-reported sleep quality (Akerstedt et al., 1997; Della Monica et al., 2018; Kaplan, Hardas, et al., 2017; Kaplan, Hirshman, et al., 2017; Keklund & Akerstedt, 1997; Svetnik et al., 2020). However, prediction models only account for up to 20% of the variance across different indices of self-reported sleep quality, suggesting that other factors are implicated in the morning perception of sleep. One factor that has received less attention is the level of autonomic arousal during sleep, which could be important for the overall restorative properties of sleep (de Zambotti, Trinder, et al., 2018) and therefore influence sleep quality assessments. For example, a study found that heart rate variability (HRV) during sleep was significantly lower in patients with chronic fatigue syndrome than in controls, and a lower HRV predicted a lower subjective sleep quality (Burton et al., 2010).

Most of the insights about the interplay between objective and subjective sleep assessments have come from studies in adults, with few studies examining the biological basis of the subjective sleep experience in adolescents. Critically, adolescence is a time of profound developmental changes across several biological domains (Blakemore & Choudhury, 2006), including physiological sleep processes (Feinberg & Campbell, 2010), sleep preferences (i.e., preferences for evening activities and later bedtimes) and specific psychosocial changes in lifestyle habits, often leading to inadequate and disrupted sleep patterns (Owens et al., 2014). In addition, the prevalence of sub-clinical sleep complaints and clinical insomnia increases during adolescence, particularly in older girls (de Zambotti, Goldstone, et al., 2018). However, contrasting physiological sleep data suggests that, if sex differences in sleep do exist in adolescence, adolescent girls tend to sleep better than boys (i.e., less PSG-measured wakefulness after sleep onset and fewer awakenings) (Baker et al., 2016).

In this study, we aimed to investigate age- and sex-dependent differences in morning sleep perception, indices of morning mood (e.g., sadness, stress, irritability), and readiness (e.g., concentration, fatigue, readiness), and the discrepancies between objective-subjective sleep in adolescents. We also investigated the physiological correlates of morning sleep perception by analyzing the relationship between overnight physiological sleep (single-night PSG sleep macrostructure and EEG sleep microstructure and sleep-autonomic nervous system functioning measures) and the next-morning perception of sleep, mood, and readiness. This study was performed on a sample of healthy adolescent boys and girls without insomnia, who were participating in the US National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA) study. We hypothesized that sex differences in the discrepancies between objective and subjective sleep outcomes would be significant in older adolescents, toward a greater overestimation of indices of sleep disruption in girls, compared to boys. We also hypothesized that sleep autonomic features would significantly contribute, over and above indices of sleep macro- and micro-structure, in explaining morning perception of sleep and related indices of mood and readiness.

2. Method

2.1. Participants

One-hundred-thirty-seven healthy adolescents (12-21 years old; 105 Caucasian) from the baseline NCANDA sleep sub-study comprised the final sample. The sample included 61 girls (Age, mean ± SD: 15.6 ± 2.3 y; Body Mass Index [BMI], mean ± SD, 21.6 ± 5.0 kg.m−2) and 76 boys (Age, mean ± SD: 15.3 ± 2.1 y; BMI, mean ± SD: 21.8 ± 4.5 kg.m−2). Details about the NCANDA study are described elsewhere (Baker et al., 2016; Brown et al., 2015).

Adolescents were recruited from SRI International (N = 109) and the University of Pittsburgh (N = 28) between 2013 and 2014, through local schools, colleges, catchment-area calling, or public notices. Data about sample characteristics are from the data release NCANDA_DATA_00010_V2 distributed to the public according to the NCANDA Data Distribution agreement: https://www.niaaa.nih.gov/national-consortium-alcohol-and-neurodevelopment-adolescence-ncanda. Briefly, all participants had a phone interview and in-person screening session including the Semi-Structured Assessment for the Genetics of Alcoholism (Bucholz et al., 1994). Exclusion criteria were minimal and centered on factors that may confound detecting main effects of interest or that compromised completion of the protocol (e.g., no current use of medications affecting brain function, no serious medical conditions). None of the participants showed evidence of sleep-disordered breathing, periodic limb movement disorder, and/or narcolepsy, as assessed by a clinical sleep evaluation. Girls who were post-menarche were studied irrespective of the menstrual cycle phase.

The study was approved by the Institutional Review Boards at SRI International and the University of Pittsburgh. Adult participants provided written consent to participate, and minors provided written assent in addition to consent from a parent/legal guardian.

2.2. Procedure

All but 8 participants had a clinical PSG screening/adaptation night, followed by one experimental PSG recording, analyzed here. At both sites, PSG recordings were carried out in temperature-controlled and sound-attenuated bedrooms. Lights were turned out according to the participants’ self-reported typical bedtimes, determined by a single retrospective “usual bedtime” question. In the morning, participants woke up at their desired time. Upon awakening, participants completed questions assessing their perceived sleep quality, mood, and readiness. Each night, a breath alcohol test (S75 Pro, BACtrack Breathalyzers, San Francisco, CA, USA) and urine drug test (10 Panel iCup drug test kit, Instant Technologies, Inc.) confirmed the absence of recent alcohol or drug use.

2.3. PSG sleep assessment

Standard PSG was performed using Compumedics Grael HD-PSG systems (Compumedics, Abbotsford, Victoria, Australia) according to American Academy of Sleep Medicine (AASM) guidelines (Iber, 2007), including the recording of EEG (F3/4, C3/4, O1/2 referred to the contralateral mastoids), submental electromyography, bipolar electrooculography, and electrocardiography (ECG).

Arousals (≤15 s) and 30-s sleep epochs (wake, N1, N2, N3, rapid-eye-movement [REM]) were manually scored according to the AASM rules. The following standard PSG indices were calculated: lights-out time (LO; hh:mm:ss); time in bed (TIB, min; calculated as the time from LO to lights-on); total sleep time (TST, min); sleep efficiency (SE, %; calculated as TST/TIB*100); sleep onset latency (SOL, min); REM latency (REML, min); wake after sleep onset (WASO, min); the amount of time (%) spent in N1, N2, N3, and REM as a percentage of TST. The arousal index (ARI) was calculated as the total number of arousals per hour of sleep and the awakening index (AWKI) was calculated as the total number of awakenings per hour of sleep.

2.4. Power spectral analysis of sleep EEG

EEG power (μV2/Hz) at a frontal site (F3) during N2 + N3 sleep was calculated for delta (0.3–4 Hz), theta (4–8 Hz), sigma (12–15 Hz), alpha (8–12 Hz), beta1 (15–23 Hz) and beta2 (23–30 Hz) bands. EEG data were collected at 256 Hz. Briefly, EEG was re-referenced to the average mastoid and filtered at 0.3–36 Hz with a half-amplitude cut-off at 0.15 and 36.15 Hz to filter out low frequencies (e.g., due to contamination like sweating) and high frequencies (e.g., due to electromagnetic interference). Fast Fourier transform analysis was conducted on each 30-s epoch using a 4-s sliding Hanning window to calculate power density values with 0.125 Hz resolution. Epochs containing arousals were excluded (no more than 5% of epochs were rejected). Please refer to (Goldstone et al., 2019) for details.

2.5. Frequency-domain HRV analysis to assess sleep ANS activity.

ECG data were collected at 1024 Hz using Ag/AgCl electrodes placed in a modified lead II Einthoven configuration, using a dedicated channel of the Compumedics Grael HD-PSG systems (Compumedics, Abbotsford, Victoria, Australia). Normal-to-normal (N-N) inter-beat intervals (ms) were calculated from the automatic detection of ECG R-waves and manually checked. According to modified rules described in Trinder et al., (Trinder et al., 2001) frequency domain HRV analysis was performed on artifact-free, 2-minute bins of undisturbed non-REM (NREM; N2 + N3) and REM sleep selected throughout the night. The two-minute bins had to be preceded by four epochs of the same sleep stage. The following indices were computed for both NREM and REM: heart rate (HR; bpm), peak frequency in the HF range (HFpf; Hz; as a measure of respiratory frequency), high frequency (HF; ms2), and low frequency (LF; ms2) narrow absolute(de Zambotti, Trinder, et al., 2018; Trinder et al., 2001) and high frequency normalized unit (HFnu; %; calculated as HF/(LF+HF). Differences in HR, HFpf, and HFnu between REM and NREM sleep were also calculated (ΔHR; ΔHFpf; ΔHFnu) as measures of ANS reactivity to sleep state shifting (de Zambotti, Trinder, et al., 2018).

2.6. Assessment of morning perception of sleep, mood, and readiness

Upon awakening, participants reported the time in minutes it took them to fall asleep (SOLsub), the number of awakenings during the night (AWKsub), and the time in minutes spent awake after falling asleep (WASOsub). In addition, they used 100 mm visual analog scales ranging from minimum to maximum intensity to rate their perceived sleep quality, depth, and restlessness as well as their current mood states (i.e., sad, tense, happy, weary, calm, anxious, exhausted, relaxed, stressed, and irritable) and readiness (i.e., able to concentrate, alert, energetic, how much of an effort it is to do things, sleepiness, the overall feeling of wellbeing, clear-headedness, fatigue, forgetfulness, and efficiency) (Buysse et al., 2007).

2.7. Statistical analyses

Discrepancies between objective and subjective SOL (ΔSOL) and WASO (ΔWASO) were calculated as the absolute difference between PSG and equivalent subjective morning measures.

Objective sleep measurements were examined for skewness, and log transformations were applied to improve normality. EEG delta, theta, alpha, sigma, beta1, and beta2 activities, PSG LO, TST, SE, SOL, WASO, ARI, AWKI, and HRV NREM and REM LF and HF were log-transformed before analyses. Outliers for both subjective and objective measurements were identified and removed when exceeding 3 standard deviations above or below the mean (between 1 and 9 cases were removed). In no regression were there more than 9 participants excluded due to missing values or outliers in the objective or subjective variables. All objective measurements other than sex and age were standardized to have zero mean and unit standard deviation. Sex was coded as 0 for males and 1 for females. Age was centered at zero but not standardized to have unit variance.

A first set of models was used to investigate age and sex differences in morning perception of sleep, mood, readiness, and on discrepancy variables ΔSOL and ΔWASO. Linear regressions were performed for any subjective measure as the dependent variable and the independent variables were sex, age, and their interaction with each other (a total of 28 models were run). Robust variance estimators (which do not require normality of residuals and are robust to heteroscedasticity) were used to determine the statistical significance of the dependent variables.

Analyses were then conducted to evaluate the prediction of the perception of sleep, mood, and readiness based on overnight PSG sleep macrostructure, sleep EEG and sleep ANS measures (a total of 26 models were run).

Lasso predictor selection was conducted with 10-fold cross-validation where the dependent variable was a subjective measure, and the independent variables were all of the PSG (LO, TST, SE, SOL, WASO, ARI, AWKI, REM, N1, N2, N3), EEG (delta, theta, sigma, alpha, beta1, and beta2), HRV NREM measures (HR, HFpf, HF, LF, HFnu), HRV REM-NREM discrepancy variables (ΔHR, ΔHFpf, ΔHFnu), sex, age, the interaction of sex and age. For each model, the selected predictors were then entered into a linear regression with robust variance estimates (Huber/White/sandwich estimators) to identify the statistically significant predictors. The change in R-squared when a selected predictor was removed from the regression equation was used as a measure of the relative importance of that predictor.

F, p, and R2 values are provided for all significant models. In addition, Benjamini-Hochberg false discovery rates (denoted "q") were calculated for p-values for the first set of 26 models and separately for the second set of 28 models to adjust for multiple comparisons. Effects were considered significant at q < 0.05. Analyses were performed using Stata/SE 16.1 for Windows by a senior biostatistician (HJ).

3. Results

3.1. Age and sex differences in the observed estimates of morning perception of sleep, mood, and readiness, and the discrepancy between PSG and perceived sleep measures.

Models were significant for predicting perceived restless sleep (F3,132 = 8.08, p < .001, R2 = 0.146, q < .01), depth of sleep (F3,132 = 4.19, p < .01, R2 = 0.066, q < .05) and AWKsub (F3,132 = 5.34, p < .01, R2 = 0.085, q < .05), with age as the significant covariate (p < 0.001, q < .05). For both girls and boys, older adolescents reported more awakenings and perceived their sleep to be deeper and less restless than younger adolescents (Figure 1). There were no significant main effects for sex or interaction between sex and age.

Figure 1.

Figure 1.

Open in a new tab

Morning perception of restless, sleep depth, and night-time awakenings, reported after a night of sleep in the laboratory, as a function of age, separately in boys (blue circles; N=76) and girls (purple triangles; N=61).

While sex differences in objective and subjective WASO discrepancies were pronounced in older adolescent boys compared to girls (see Figure 2), the model predicting ΔWASO was no longer significant after adjusting for multiple comparisons. The model predicting ΔSOL was also not significant.

Figure 2.

Figure 2.

Open in a new tab

Discrepancies in objective-subjective wake after sleep onset (WASO), as a function of age, separately in boys (blue circles; N=76) and girls (purple triangles; N=61).

3.2. Prediction of morning sleep perception based on overnight PSG sleep macrostructure, EEG sleep microstructure, and sleep ANS measures.

Models that included selected predictors (see Figure 3) among night-time PSG, EEG, and ANS measures were significant in predicting the morning perception of sleep (SOLsub, AWKsub, WASOsub, sleep quality, restless and depth of sleep), mood (sad, exhausted, relaxed, stressed, irritability) and readiness (alert, fatigued) variables. Models explained between 3% and 29% of the variance as measured by R2 (see Figure 3 for F, p, q, and R2-values for all significant models).

Figure 3.

Figure 3.

Open in a new tab

Significant models and related demographics, polysomnographic (PSG), electroencephalographic (EEG), and autonomic features, and their relative contribution (%) in explaining the morning perception of sleep, mood, and readiness. Directionality in the relationship between features and the target morning variable is also highlighted (+/−). LO, lights-off; TST, total sleep time; SE, sleep efficiency; SOL, sleep onset latency; WASO, wake after sleep onset; AWKI, awakening index; REM, rapid-eye-movement; NREM, non-REM; HF, high frequency (nu = normalized units; pf = peak frequency); HR, heart rate.

For each of the significant models, the significant factors and their relative contributions are displayed in Figure 3. Overall, indices of PSG macrostructure explained most of the variance in the significant prediction of morning sleep, mood, and readiness variables. Age and sex significantly contributed to the prediction of morning sleep perception variables but not mood and readiness variables. EEG measures mostly contributed to the prediction of morning sleep perception variables, whereas ANS indices were important predictors for sleep perception, mood, and readiness variables (Figure 3).

Specifically, ΔHFnu, reflecting the shifting in vagal dominance between NREM and REM (average drop in vagal dominance from NREM to REM, mean ± SD = −18.4 ± 10.6%) was the best predictor of subjective sleep variables. ΔHFnu was positively associated with perceived sleep depth, with a higher drop in vagal dominance from NREM to REM being associated with deeper sleep. ΔHFnu was negatively associated with perceived restless sleep, with a smaller drop in vagal dominance from NREM to REM being associated with a more restless sleep (p < 0.05). In addition, ΔHFnu was positively related to more perceived stress, with a greater NREM-REM shift in vagal dominance associated with more perceived stress (p < 0.05).

On the other hand, HR during NREM was the best predictor of the morning mood variables. HR during NREM was negatively associated with and contributed to predicting feeling more exhausted and irritable. It was also positively associated with and contributed to predicting alertness (p < 0.05) and was negatively related and contributed to predicting fatigue. Finally, HFpf (reflecting respiratory frequency) during NREM was negatively associated with and contributed to predicting stress (p<0.05), with individuals who had lower breathing rates during NREM sleep perceiving higher levels of stress in the morning.

4. Discussion

Our findings in a large sample of healthy adolescents indicate that perception of sleep differs according to age, with older adolescents reporting that their sleep was deeper and less restless, although they reported more nocturnal awakenings than younger adolescents. However, our data did not support the hypothesis that older girls overestimate the extent of their objective sleep disruption compared with older boys. On the contrary, albeit not significant, older adolescent boys tended to underestimate the objective amount of total wake time at night. To our knowledge, only one study investigated sex and age differences in objective-subjective sleep discrepancies in adolescents, showing that actigraphy WASO was underestimated, while TST was overestimated, and greater objective-subjective discrepancies were found in boys compared to girls (Short et al., 2012). Further studies are needed to further investigate biological sex as a relevant factor in explaining objective-subjective sleep discrepancies.

The perception of sleep is critical in the context of insomnia, a common sleep disorder, which becomes more prevalent in girls than boys, post-puberty (de Zambotti, Goldstone, et al., 2018). Our results suggest that subjective-objective discrepancies in WASO do not significantly change across age, although it tended to be larger in older adolescent boys. A better understanding of these discrepancies could help to elucidate the contrasting literature indicating that, as they get older, the incidence of insomnia is greater in girls compared to boys, despite overall indication of no differences or better objective sleep in girls, compared to boys. Longitudinal follow-ups of objective-subjective differences in sleep in adolescents at risk for developing insomnia may ultimately confirm this hypothesis.

As hypothesized, sleep autonomic features significantly contributed, together with indices of sleep macro- and micro-structure, to the perception of sleep, mood, and readiness indices. Overall, our models explained between 3% and 29% of the between-subject variance in morning subjective sleep quality, mood, and readiness scores, with more than 70% of the variance remaining unexplained. In agreement with previous studies (Della Monica et al., 2018; Kaplan, Hardas, et al., 2017; Kaplan, Hirshman, et al., 2017; Owens et al., 2010), PSG measures of sleep duration and continuity appeared to be the strongest predictors of morning sleep perception, mood, and readiness. Furthermore, similar to our finding, in a sample that included young adults (206 men and women, aged 20–84 years, without sleep complaints), Della Monica et al. (Della Monica et al., 2018) found that sleep quality was negatively associated with PSG number of awakenings and positively associated with the duration of REM sleep. They also found that the associations between objective (latency to persistent sleep, TST, SE, number of awakenings, Stage 1 sleep duration) and subjective sleep quality were stronger in women than in men.

An additional finding from our study is that a lower TST predicted higher levels of sadness, stress, exhaustion, and irritability, and reduced alertness, supporting the importance of achieving a proper sleep amount for optimal morning mood and readiness. For example, those adolescents sleeping less than 7 hours (30.7% of the total sample) had, on average, 7.7% reduced morning alertness compared to those achieving at least 7 hours of sleep. This finding supports numerous studies confirming the relationship between sleep, mood, fatigue, and morning performance, with insufficient sleep being related to lack of motivation, inattention, poor decision-making skills (Harrison & Horne, 2000), and daytime sleepiness (O'Brien & Mindell, 2005), impacting academic performance (Dewald et al., 2010) and affecting executive function (Beebe, 2011) in adolescents.

While most prior studies focused on traditional sleep measures as predictors of morning self-reported sleep and related measures, we also included indices of ANS functioning in our study. Our results indicate that ANS measures contributed to predicting subjective sleep, mood, and readiness ratings. In contrast, Faerman and colleagues (Faerman et al., 2020) found that ANS factors were not significant predictors of subjective sleep quality in a sample of 1141 elderly men. Discrepancies in findings could be attributed to the vast difference in age between samples as well as in the specific ANS measures used (only overnight averages of HR and HRV were used by Faerman and colleagues). In our analysis, the ANS measure reflecting NREM-REM shifting in vagal activity was the best predictor of morning self-reported sleep. For example, the perception of a deeper sleep was associated with a bigger difference in vagal dominance between NREM and REM sleep, while the perception of a more restless sleep was associated with a smaller difference in vagal dominance between NREM and REM sleep, suggesting that sleep stage-related shifts in ANS function, beyond overnight averages of overall ANS activity, are implicated in morning subjective perceptions about sleep. Overall, these findings support studies highlighting the importance of ANS recovery during sleep (de Zambotti, Trinder, et al., 2018) and the reciprocal interaction between ANS, mood, fatigue, and morning performance. The ANS drives sensory modulation in regulating and organizing reactions to sensations in a graded and adaptive manner (Ayres, 1972; Brown et al., 2019). Moreover, the ANS also affects cognitive processing, with higher HRV and slower HR being associated with better cognitive performance in adults (Colzato et al., 2018; Elias & Torres, 2017). Also, the ANS plays an important role in mediating emotion, emotion recognition (Quintana et al., 2012), cognition, and behavior (Critchley et al., 2013). Further, a large proportion of variance in sleep-dependent memory consolidation effects was shown to be explained by sleep ANS activity, on top of traditional sleep features (e.g., spindle activity) (Whitehurst et al., 2016).

In our models, respiratory frequency during sleep (as derived from HRV HF peak frequency) also contributed to subjective morning mood. In particular, a lower breathing rate during NREM sleep was associated with a higher level of perceived stress in the morning. It is unclear what drives these associations, and further work is needed, also considering the potential role of mediators such as basal cardio-fitness levels.

Since previous studies and the current study show that objective measures of sleep do not explain much variance in subjective sleep ratings, other factors may be important, including awareness about the prior night of sleep, the accuracy of recall, and how perception is influenced. It has been shown that when stimulated prior to wake, specific neuronal populations showed use-dependent patterns in sleep intensity during NREM sleep (Kattler et al., 1994). Moreover, Kay and colleagues (Kay et al., 2017) investigated the associations between SOL discrepancy and relative regional cerebral metabolic rate for glucose (rCMRglc, measure of regional brain activity) during NREM sleep and showed that, in primary insomnia patients, larger SOL discrepancies (self-reported > PSG-measured SOL) were associated with significantly higher relative rCMRglc in brain networks involved in self-referential thinking or awareness. A greater discrepancy in the time spent falling asleep (SOL) was also associated with a cortical EEG index of hyperarousal across the sleep onset period (Maes et al., 2014) providing new insights into the physiological processes underlying the subjective sleep experience. These results highlight the potential influence of awareness factors and the possibility that local variations of sleep intensity may be implicated in objective-subjective discrepancies.

This study has several limitations that should be noted. The study is observational and cross-sectional and does not allow the assessment of causality. The analyses were performed on limited data with the risk of overfitting models, and study replication should be performed in independent datasets to improve generalization. Also, the in-lab sleep assessment may not reflect the participants’ typical sleep. Laboratory experimental manipulations (e.g., selective sleep deprivation) and more refined analytics depicting multi-system relationships during the night (e.g., cortical-cardiac coupling) may further advance our understanding of the specific sleep process implicated in the perception of sleep. The use of a single night is also a limitation since we could only infer between-person variance and could not consider within-person variance. A large-scale longitudinal evaluation of multiple nights of sleep would be preferable to evaluate the extent of within-person variation of objective sleep measures and the extent to which they explain the next-morning sleep perception. The use of multi-sensor wearable technology paired with ecologic momentary assessments could help to model between- and within-person variation and advance our understanding of factors (e.g., demographics, psychophysiological state, disease condition, substance use, school time vs vacation time, week vs weekend) implicated in the objective-subjective sleep relationship in adolescents in the real world. Future investigations should also consider psychosocial factors. For example, it has been suggested that underestimation of TST is not a generic characteristic of insomnia, but rather that personality traits and perhaps even constitutional factors seem to influence the perception of sleep duration and wakefulness (Means, 2003).

In summary, the subjective experience of sleep and sleep-related mood and readiness differs according to age in an adolescent sample and is a complex construct with multiple components (biopsychosocial, demographics, and other individual factors) that need to be considered when characterizing the subjective-objective sleep dynamics.

Financial disclosure.

This work was made possible with support from the National Institutes of Health, the National Consortium on Alcohol and NeuroDevelopment in Adolescence (NCANDA) grants AA021696 (FCB, MdZ), and AA021690 (DBC). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Non-financial disclosure. The authors declared no conflict of interest related to the current work. MdZ and FCB have received research funding unrelated to this work from Noctrix Health, Inc., and Lisa Health Inc. MdZ is a co-founder of Lisa Health Inc. MdZ and FCB have ownership of shares in Lisa Health.

Data Availability Statement

The data underlying this article will be shared on reasonable request to the corresponding author.

References

  1. Akerstedt T, Hume K, Minors D, & Waterhouse J (1997, Dec). Good sleep--its timing and physiological sleep characteristics. J Sleep Res, 6(4), 221–229. 10.1111/j.1365-2869.1997.00221.x [DOI] [PubMed] [Google Scholar]
  2. Ayres AJ (1972). Improving academic scoresthrough sensory integration. Journal of Learning Disabilities, 5(6), 338–343. [Google Scholar]
  3. Baker FC, Willoughby AR, de Zambotti M, Franzen PL, Prouty D, Javitz H, Hasler B, Clark DB, & Colrain IM (2016, Jul 1). Age-Related Differences in Sleep Architecture and Electroencephalogram in Adolescents in the National Consortium on Alcohol and Neurodevelopment in Adolescence Sample. Sleep, 39(7), 1429–1439. 10.5665/sleep.5978 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Beebe DW (2011, Jun). Cognitive, behavioral, and functional consequences of inadequate sleep in children and adolescents. Pediatr Clin North Am, 58(3), 649–665. 10.1016/j.pcl.2011.03.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Blakemore SJ, & Choudhury S (2006, Mar-Apr). Development of the adolescent brain: implications for executive function and social cognition. J Child Psychol Psychiatry, 47(3-4), 296–312. 10.1111/j.1469-7610.2006.01611.x [DOI] [PubMed] [Google Scholar]
  6. Brown A, Tse T, & Fortune T (2019, Dec). Defining sensory modulation: A review of the concept and a contemporary definition for application by occupational therapists. Scand J Occup Ther, 26(7), 515–523. 10.1080/11038128.2018.1509370 [DOI] [PubMed] [Google Scholar]
  7. Brown SA, Brumback T, Tomlinson K, Cummins K, Thompson WK, Nagel BJ, De Bellis MD, Hooper SR, Clark DB, Chung T, Hasler BP, Colrain IM, Baker FC, Prouty D, Pfefferbaum A, Sullivan EV, Pohl KM, Rohlfing T, Nichols BN, Chu W, & Tapert SF (2015, Nov). The National Consortium on Alcohol and NeuroDevelopment in Adolescence (NCANDA): A Multisite Study of Adolescent Development and Substance Use. J Stud Alcohol Drugs, 76(6), 895–908. 10.15288/jsad.2015.76.895 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bucholz KK, Cadoret R, Cloninger CR, Dinwiddie SH, Hesselbrock VM, Nurnberger JI Jr., Reich T, Schmidt I, & Schuckit MA (1994, Mar). A new, semi-structured psychiatric interview for use in genetic linkage studies: a report on the reliability of the SSAGA. J Stud Alcohol, 55(2), 149–158. 10.15288/jsa.1994.55.149 [DOI] [PubMed] [Google Scholar]
  9. Burton AR, Rahman K, Kadota Y, Lloyd A, & Vollmer-Conna U (2010, Jul). Reduced heart rate variability predicts poor sleep quality in a case-control study of chronic fatigue syndrome. Exp Brain Res, 204(1), 71–78. 10.1007/s00221-010-2296-1 [DOI] [PubMed] [Google Scholar]
  10. Buysse DJ, Thompson W, Scott J, Franzen PL, Germain A, Hall M, Moul DE, Nofzinger EA, & Kupfer DJ (2007, Apr). Daytime symptoms in primary insomnia: a prospective analysis using ecological momentary assessment. Sleep Med, 8(3), 198–208. 10.1016/j.sleep.2006.10.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Colzato LS, Jongkees BJ, de Wit M, van der Molen MJW, & Steenbergen L (2018, Aug). Variable heart rate and a flexible mind: Higher resting-state heart rate variability predicts better task-switching. Cogn Affect Behav Neurosci, 18(4), 730–738. 10.3758/s13415-018-0600-x [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  12. Critchley HD, Eccles J, & Garfinkel SN (2013). Interaction between cognition, emotion, and the autonomic nervous system. Handb Clin Neurol, 117, 59–77. 10.1016/b978-0-444-53491-0.00006-7 [DOI] [PubMed] [Google Scholar]
  13. de Zambotti M, Goldstone A, Colrain IM, & Baker FC (2018, Jun). Insomnia disorder in adolescence: Diagnosis, impact, and treatment [Review]. Sleep Med Rev, 39, 12–24. 10.1016/j.smrv.2017.06.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. de Zambotti M, Trinder J, Silvani A, Colrain IM, & Baker FC (2018, Jul). Dynamic coupling between the central and autonomic nervous systems during sleep: A review [Review]. Neurosci Biobehav Rev, 90, 84–103. 10.1016/j.neubiorev.2018.03.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Della Monica C, Johnsen S, Atzori G, Groeger JA, & Dijk DJ (2018). Rapid Eye Movement Sleep, Sleep Continuity and Slow Wave Sleep as Predictors of Cognition, Mood, and Subjective Sleep Quality in Healthy Men and Women, Aged 20-84 Years. Front Psychiatry, 9, 255. 10.3389/fpsyt.2018.00255 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dewald JF, Meijer AM, Oort FJ, Kerkhof GA, & Bogels SM (2010, Jun). The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: A meta-analytic review. Sleep Med Rev, 14(3), 179–189. 10.1016/j.smrv.2009.10.004 [DOI] [PubMed] [Google Scholar]
  17. Edinger JD, & Krystal AD (2003, Jun). Subtyping primary insomnia: is sleep state misperception a distinct clinical entity? Sleep Med Rev, 7(3), 203–214. 10.1053/smrv.2002.0253 [DOI] [PubMed] [Google Scholar]
  18. Elias MF, & Torres RV (2017). The Renaissance of Heart Rate Variability as a Predictor of Cognitive Functioning. American Journal of Hypertension, 31(1), 21–23. 10.1093/ajh/hpx150 [DOI] [PubMed] [Google Scholar]
  19. Faerman A, Kaplan KA, & Zeitzer JM (2020, Sep). Subjective sleep quality is poorly associated with actigraphy and heart rate measures in community-dwelling older men. Sleep Med, 73, 154–161. 10.1016/j.sleep.2020.04.012 [DOI] [PubMed] [Google Scholar]
  20. Feinberg I, & Campbell IG (2010, Feb). Sleep EEG changes during adolescence: an index of a fundamental brain reorganization. Brain Cogn, 72(1), 56–65. 10.1016/j.bandc.2009.09.008 [DOI] [PubMed] [Google Scholar]
  21. Goldstone A, Willoughby AR, de Zambotti M, Clark DB, Sullivan EV, Hasler BP, Franzen PL, Prouty DE, Colrain IM, & Baker FC (2019, Jun). Sleep spindle characteristics in adolescents. Clin Neurophysiol, 130(6), 893–902. 10.1016/j.clinph.2019.02.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Harrison Y, & Horne JA (2000, Sep). The impact of sleep deprivation on decision making: a review. J Exp Psychol Appl, 6(3), 236–249. 10.1037//1076-898x.6.3.236 [DOI] [PubMed] [Google Scholar]
  23. Harvey AG, & Tang NK (2012, Jan). (Mis)perception of sleep in insomnia: a puzzle and a resolution. Psychol Bull, 138(1), 77–101. 10.1037/a0025730 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Iber C A. A. o. S. M. (2007). The AASM manual for the scoring of sleep and associated events : rules, terminology and technical specifications. American Academy of Sleep Medicine. [Google Scholar]
  25. Kaplan KA, Hardas PP, Redline S, Zeitzer JM, & Sleep Heart Health Study Research, G. (2017, Jun). Correlates of sleep quality in midlife and beyond: a machine learning analysis. Sleep Med, 34, 162–167. 10.1016/j.sleep.2017.03.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kaplan KA, Hirshman J, Hernandez B, Stefanick ML, Hoffman AR, Redline S, Ancoli-Israel S, Stone K, Friedman L, Zeitzer JM, & Osteoporotic Fractures in Men, S. o. O. F. S. O. F. R. G. (2017, Feb). When a gold standard isn't so golden: Lack of prediction of subjective sleep quality from sleep polysomnography. Biol Psychol, 123, 37–46. 10.1016/j.biopsycho.2016.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kattler H, Dijk DJ, & Borbély AA (1994, Sep). Effect of unilateral somatosensory stimulation prior to sleep on the sleep EEG in humans. J Sleep Res, 3(3), 159–164. 10.1111/j.1365-2869.1994.tb00123.x [DOI] [PubMed] [Google Scholar]
  28. Kay DB, Karim HT, Soehner AM, Hasler BP, James JA, Germain A, Hall MH, Franzen PL, Price JC, Nofzinger EA, & Buysse DJ (2017, Nov 1). Subjective-Objective Sleep Discrepancy Is Associated With Alterations in Regional Glucose Metabolism in Patients With Insomnia and Good Sleeper Controls. Sleep, 40(11). 10.1093/sleep/zsx155 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Keklund G, & Akerstedt T (1997, Dec). Objective components of individual differences in subjective sleep quality. J Sleep Res, 6(4), 217–220. 10.1111/j.1365-2869.1997.00217.x [DOI] [PubMed] [Google Scholar]
  30. Krystal AD, & Edinger JD (2008). Measuring sleep quality. Sleep Medicine, 9, S10–S17. 10.1016/s1389-9457(08)70011-x [DOI] [PubMed] [Google Scholar]
  31. Maes J, Verbraecken J, Willemen M, De Volder I, van Gastel A, Michiels N, Verbeek I, Vandekerckhove M, Wuyts J, Haex B, Willemen T, Exadaktylos V, Bulckaert A, & Cluydts R (2014, Mar). Sleep misperception, EEG characteristics and autonomic nervous system activity in primary insomnia: a retrospective study on polysomnographic data. Int J Psychophysiol, 91(3), 163–171. 10.1016/j.ijpsycho.2013.10.012 [DOI] [PubMed] [Google Scholar]
  32. Means M (2003). Accuracy of sleep perceptions among insomnia sufferers and normal sleepers. Sleep Medicine, 4(4), 285–296. 10.1016/s1389-9457(03)00057-1 [DOI] [PubMed] [Google Scholar]
  33. O'Brien EM, & Mindell JA (2005). Sleep and risk-taking behavior in adolescents. Behav Sleep Med, 3(3), 113–133. 10.1207/s15402010bsm0303_1 [DOI] [PubMed] [Google Scholar]
  34. Owens J, Adolescent Sleep Working, G., & Committee on, A. (2014, Sep). Insufficient sleep in adolescents and young adults: an update on causes and consequences. Pediatrics, 134(3), e921–932. 10.1542/peds.2014-1696 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Owens JA, Belon K, & Moss P (2010, Jul). Impact of delaying school start time on adolescent sleep, mood, and behavior. Arch Pediatr Adolesc Med, 164(7), 608–614. 10.1001/archpediatrics.2010.96 [DOI] [PubMed] [Google Scholar]
  36. Quintana DS, Guastella AJ, Outhred T, Hickie IB, & Kemp AH (2012, Nov). Heart rate variability is associated with emotion recognition: direct evidence for a relationship between the autonomic nervous system and social cognition. Int J Psychophysiol, 86(2), 168–172. 10.1016/j.ijpsycho.2012.08.012 [DOI] [PubMed] [Google Scholar]
  37. Short MA, Gradisar M, Lack LC, Wright H, & Carskadon MA (2012, Apr). The discrepancy between actigraphic and sleep diary measures of sleep in adolescents. Sleep Med, 13(4), 378–384. 10.1016/j.sleep.2011.11.005 [DOI] [PubMed] [Google Scholar]
  38. Svetnik V, Snyder ES, Tao P, Roth T, Lines C, & Herring WJ (2020, Mar). How well can a large number of polysomnography sleep measures predict subjective sleep quality in insomnia patients? Sleep Med, 67, 137–146. 10.1016/j.sleep.2019.08.020 [DOI] [PubMed] [Google Scholar]
  39. Trinder J, Kleiman J, Carrington M, Smith S, Breen S, Tan N, & Kim Y (2001, Dec). Autonomic activity during human sleep as a function of time and sleep stage. J Sleep Res, 10(4), 253–264. 10.1046/j.1365-2869.2001.00263.x [DOI] [PubMed] [Google Scholar]
  40. Whitehurst LN, Cellini N, McDevitt EA, Duggan KA, & Mednick SC (2016, Jun 28). Autonomic activity during sleep predicts memory consolidation in humans. Proc Natl Acad Sci U S A, 113(26), 7272–7277. 10.1073/pnas.1518202113 [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

The data underlying this article will be shared on reasonable request to the corresponding author.

RESOURCES