The Negative Effects of Travel on Student Athletes Through Sleep and Circadian Disruption

Collegiate athletes must satisfy the academic obligations common to all undergraduates, but they have the additional structural and social stressors of extensive practice time, competition schedules, and frequent travel away from their home campus. Clearly such stressors can have negative impacts on both their academic and athletic performances as well as on their health. These concerns are made more acute by recent proposals and decisions to reorganize major collegiate athletic conferences. These rearrangements will require more multi-day travel that interferes with the academic work and personal schedules of athletes. Of particular concern is additional east-west travel that results in circadian rhythm disruptions commonly called jet lag that contribute to the loss of amount as well as quality of sleep. Circadian misalignment and sleep deprivation and/or sleep disturbances have profound effects on physical and mental health and performance. We, as concerned scientists and physicians with relevant expertise, developed this white paper to raise awareness of these challenges to the wellbeing of our student-athletes and their co-travelers. We also offer practical steps to mitigate the negative consequences of collegiate travel schedules. We discuss the importance of bedtime protocols, the availability of early afternoon naps, and adherence to scheduled lighting exposure protocols before, during, and after travel, with support from wearables and apps. We call upon departments of athletics to engage with sleep and circadian experts to advise and help design tailored implementation of these mitigating practices that could contribute to the current and long-term health and wellbeing of their students and their staff members.

0198 Remote Saliva Sample Collection for Dim Light Melatonin Onset (DLMO) Measurement in Urban Children with Asthma During the COVID-19 Pandemic

Introduction

The COVID-19 pandemic has challenged researchers to use remote data collection. Our project includes determining DLMO phase, requiring a family-friendly without face-to-face interaction. We describe here our protocol, experiences, lessons learned, and findings from the first 15 participants.

Methods

Fifteen urban-dwelling children with moderate to severe persistent asthma [7 girls, ages 7 (n=1) to 10 years; and 8 boys, 8 or 9 years] and caregiver (CG) participated. CG tracked bedtimes and risetimes in daily diaries for 10-14 days; average bedtimes from 5 nights preceding saliva collection were used to determine timing for 10 half-hourly samples. CG and child were oriented and then watched a demo video. A “spit-kit” was delivered to the home the afternoon of the study. Kits included a small cooler bag with bottle of water, 10 numbered and 5 spare Salivette tubes (Starstedt, Germany), plastic bag, dark wraparound glasses with securing strap, and log sheet. Data collection began with a zoom call with staff, CG, and child to reiterate the instructions, answer questions, and observe the first sample. Thereafter, a staff member telephoned the caregiver every 30 minutes to prompt the next sample and query whether glasses had been kept on. CG placed kit outside the home for morning pick up. Samples were centrifuged and frozen (-20°) until sending to the assay lab (SolidPhase, Portland, ME) for melatonin radioimmunoassay (Alpco, Windham, NH).

Results

DLMO phase was determined with a 4pg/ml threshold for 11 children. DLMO phases (mtime=21:46±68 min) and average bedtimes (mtime=20:40±88min) were positively correlated (r=.87). Challenges identified for missed DLMOs included: one child supervised by a teenaged sibling (not CG); one child/CG identified as potentially uncooperative. The other two “misses” likely arose from low saliva quantity, inconsistencies with staff training, and inadequate description of requirements for wearing glasses. Procedure modifications included strategies tailored to families’ needs, experiences, and home environment that can challenge adherence to protocol, greater emphasis on wearing glasses, and cartoon reminder card and scales added to kit. Subsequent samples were successful.

Conclusion

Our approach was effective for determining DLMO phase in children using a remote approach with careful application of methods.

Support (If Any)

R01HL142058, P20GM139743

0221 Pre-Sleep Breath alcohol concentrations (psBrAC) and sleep polysomnography (PSG)

Introduction

Alcohol use before bedtime has been shown to alter sleep including decreasing sleep latency, decreasing sleep efficiency, and fragmenting sleep stage distribution. Few studies manipulate pre-sleep alcohol concentration, instead focusing on a target dose or peak alcohol concentration during the night. Thus, we investigated how presleep breath alcohol concentrations (psBrAC) level (targeting a BrAC of 0.08), are associated with same-night sleep.

Methods

Thirty (15F; ages=22-57, mean=33yr) healthy adults who self-reported moderate drinking were instructed to maintain a consistent sleep schedule (8-9h time in bed) for a least 7-days before entering a cross-over design involving two sets (separated by>3 days) of 3 consecutive nights of in-lab polysomnography. For all nights in each condition, participants drank either mixer alone or mixer+alcohol in 3 portions across 45 minutes ending 1h before lights out. psBrAC was measured within 5 minutes of lights out. PSGs were staged according to Rechtschaffen and Kales (1968). We computed: slow wave sleep (SWS) and REM as percentages of total sleep for the full night (%SWS) & (%REM), %SWS in the first third of the night (%SWST1), %REM in the last third of the night (%REMT3), minutes until sleep onset from lights out (Sleep Latency), and minutes after sleep onset until REM onset. Linear regressions tested if psBrAC on the first mixer+alcohol night predicted changes in sleep using the average of the three nights of sleep in the mixer-only condition as a covariate.

Results

psBrAC values ranged from 0.038 to 0.087 (mean = 0.066) mg/L. We identified minimal influence of BrAC on sleep: % SWS (β = -0.78; p = 0.07), % REM (β = 0.032; p = 0.96), % SWS T1 (β = -0.52; p = 0.522), % REM T3 (β = -1.01; p = 0.125), REM latency (β = 3.95; p = 0.302), and sleep latency (β = -1.23; p = 0.13).

Conclusion

These findings indicate that psBrAC was not associated with any of our sleep variables, when adjusting for sleep on nights without alcohol. Future work will examine peak BrAC as well whether the effects observed change over multiple nights of pre-sleep alcohol consumption.

Support (If Any)

R01AA025593

0181 Characterizing Sleep Regularity from Actigraphy in Younger and Older Adolescents

Introduction

Many adolescents experience variable sleep timing and restricted duration attributable to biopsychosocial influences. The Sleep Regularity Index (SRI) captures inter-daily stability of sleep/wake intervals as the likelihood of being asleep/awake at consistent times day-to-day. The SRI may capture unique dimensions of adolescent sleep given the ability to capture highly variable sleep/wake timing (including napping); however, SRI’s relative role in maturational sleep processes remains unknown. This study characterizes the SRI and sleep correlates (bedtime, midpoint, risetime, duration, and efficiency) in younger and older adolescents, including age-based comparisons.

Methods

Cross-sectional data were drawn from two cohorts: 30 younger (ages 9-10 years; 13 female; 24 White) and 38 older (ages 15-16 years; 20 female; 26 White) adolescents. Participants provided 7 consecutive nights (M=6.93±0.36) of sleep diaries and actigraphy on a self-selected sleep schedule while attending school. SRI was calculated as the probability of being asleep/wake at two points 24-hours apart (Philips et al., 2017), with higher scores demonstrating more regular sleep across days.

Results

SRI scores and distributions were similar between younger (M=79±9, range=58-94) and older (M=80±7, range=64-91) adolescents (t[66]=-0.58, p=.56). On average, younger adolescents reported a bedtime of 21:41±31, midpoint of 02:14±30, risetime of 06:47±36, and sleep duration of 9.11±0.52 hours. In contrast, older peers reported a later bedtime of 22:46±41 (t[66]=-7.21, p<.001) and midpoint of 02:47±29 (t[66]=-4.66, p<.001), with consistent risetime 06:49±29 (t[66]=-0.17, p=.87), and thus shorter sleep duration of 8.06±0.70 hours (t[66]=6.84, p<.001). In both cohorts, SRI was correlated with less wake-time after sleep onset (rs=-.93 to -.83, ps<.001) and greater sleep efficiency (rs=.80-.93, ps<.001), but not sleep duration or timing (ps=.18-.62).

Conclusion

This adolescent sample demonstrated greater sleep/wake regularity compared to previous reports of college students and adolescents/young adults, supporting the hypothesis that SRI may be a proxy for regularity of other aspects of daily living (e.g., fixed school start times). Adolescent SRI appears to be independent of sleep duration (consistent with previous findings) and timing, suggesting that SRI captures a distinct dimension of sleep. This research team plans to proceed with longitudinal analysis to clarify developmental trends, further explicating the potential informative role of SRI.

Support (If Any)

NIH R01 AA013252 and P20 GM139743.

0226 Circadian influence on food intake among adolescents with overweight and healthy weight

Introduction

Meal timing has been linked to obesity in adults and children; however, evidence for an endogenous influence of the circadian system on food intake is unknown. We measured food intake during forced desynchrony (FD) in adolescents with healthy weight (HW) or overweight (OW), hypothesizing that circadian timing would affect food intake independent of the environmental cycle and that the food intake rhythm would be delayed in adolescents with OW compared to HW.

Methods

Participants were 51 (29m) adolescents (12-15yr); 24 were HW and 27 OW determined by CDC norms. Participants completed seven, 28h-FD cycles; 6 meals occurred at fixed times each cycle with foods selected ~1h before meals and weighed before and after. Each meal’s energy intake was computed relative to total energy consumed in that FD cycle. Endogenous circadian period was determined using salivary dim-light melatonin onset (DLMO) phases (DLMO=0°). Time awake effect was assessed using mixed effect models and circadian phase with multilevel cosinor that included time awake as a categorical covariate.

Results

Circadian period was not significantly different between OW and HW (mean, StDev; HW =24.19h,0.22; OW=24.22h,0.14). Overall, participants consumed more calories on the cycles’ first meals (22.0% [21.3;22.6]) compared to the last meal (12.4% [11.8;13.0]). Adolescents with OW vs. HW consumed a higher proportion of calories later in the wake episode (F(5,2081)=2.63,p=.02). As hypothesized, circadian phase influenced caloric intake with an amplitude of 2.66% [2.19;3.14] percent daily calories and an acrophase of 301°[291;310], equivalent to ~17:52 in this population. The circadian influence differed by weight category (likelihood ratio test of both amplitude and acrophase; χ2(2)=10.7,p<.01), with those with OW showing a lower amplitude (OW = 2.11%[1.40;2.82], HW=3.53%[2.94;4.12] and later acrophase (OW=301° [287;314], HW=290° [278;302]).

Conclusion

Results show for the first time an independent influence of the endogenous circadian timing system on caloric intake of humans that differed as a function of body weight: caloric intake for adolescents with OW had a circadian rhythm with blunted amplitude and delayed peak phase. These observations show circadian control of food intake that may be stronger in HW than OW adolescents.

Support (If Any)

P20GM139743; R01DK101046; R01HL153969

0256 Child Maltreatment and Multidimensional Sleep Health among Incoming First-Year College Students

Introduction

Despite the growing body of evidence linking child maltreatment to compromised sleep health in adulthood, links in emerging adults are understudied. We examined associations between child maltreatment (CM) and multidimensional sleep health among emerging adults undergoing the major life transition of starting college.

Methods

First-year college students (N=682, 41% male, 48% Non-Hispanic White, 22% Non-Hispanic Asian, 15% Hispanic all races, 6% Non-Hispanic Black, and 9% Non-Hispanic other races) completed daily sleep diaries (DSDs) for 9 weeks, and completed the Childhood Trauma Questionnaire (CTQ), Epworth Sleepiness Scale (ESS), and Pittsburgh Sleep Quality Index (PSQI) following DSD completion. We used linear regression models to examine associations between CTQ-derived CM (0=none, 1=any [moderate to severe emotional abuse/neglect, physical abuse/neglect, or sexual abuse]) and sleep health (Buysse, 2014) using a multidimensional score encompassing components from the RUSATED model (regularity [DSD sleep midpoint SD: 0= >1 hour, 1= ≤1 hour], satisfaction [PSQI sleep quality item: 0=fairly or very bad, 1=very or fairly good], alertness [ESS score: 0= >10, 1= ≤10], timing [DSD sleep midpoint: 0= <3:30 or >5:30, 1= ≥3:30 and ≤5:30], efficiency [DSD sleep efficiency: 0= <93%, 1= ≥93%], and duration [DSD sleep duration: 0= <7 hours or >10 hours, 1= ≥7 hours and ≤10 hours].

Results

In the full sample 20.5% reported CM (within-group prevalences: females 21%, males 20%, Non-Hispanic Whites 12%, Non-Hispanic Asians 28%, Hispanics of all races 26%, Non-Hispanic Blacks 34%, and Non-Hispanics of other races 30%). Those with CM had significantly worse sleep health (B=-0.25, 95% CI=-0.46, -0.04) compared to those without CM, but not after adjustment for sex and race/ethnicity. In logistic regression models, the only individual sleep health component significantly associated with CM was sleep satisfaction. After adjustment for sex, race/ethnicity, and depressive symptoms, those who experienced CM had a 52% lower odds of reporting good sleep quality (OR=0.48, 95% CI=0.30, 0.76).

Conclusion

CM is associated with worse sleep satisfaction among first-year college students, which aligns with previous research in older adults. Additional research should examine neurophysiological correlates of sleep satisfaction in the context of child maltreatment and effects on subsequent health.

Support (If Any)

P206M139743, MH079179, T32HD101392.

276 The Actigpatch: validation of a novel adhesive monitor against PSG and wrist-actigraphy

Introduction

Wrist actigraphy is a gold-standard method for estimating sleep patterns in the field. Actigraphy adherence is limited when participants remove the device for daily activities (e.g., showers, exercise). Here we evaluate the validity of a novel water-resistant wearable, the “Actigpatch,” compared to polysomnography and traditional actigraphy.

Methods

Seven adults (4F; aged 22-54 years [m: 31.1±13.1]) slept in the laboratory for a total of 33 nights. Participants wore a Micro Motionlogger actigraphy (Ambulatory Monitoring Inc., Ardley, NY) on the non-dominant wrist and the Actigpatch—a 0.5in2 circuit board enclosed in a water-resistant adhesive (Circadian Positioning Systems, Newport, RI)—on the triceps. Both devices recorded tri-axial accelerometry, with sleep-wake estimates produced in 1-minute epochs (Sadeh algorithm). Simultaneous PSG data were reduced to 1-minute resolution favoring wake, keeping with recent recommendations. We computed epoch-by-epoch confusion matrices and derived 2 validation parameters: sensitivity (e.g., ability to detect sleep) and specificity (e.g., ability to detect wake). Finally, we compared total sleep time estimates (TST) to evaluate the bias of each device. Nested mixed models (nights within individuals) compared device performance.

Results

The Actigpatch demonstrated high sensitivity (.95; 95%CI: [.92 .98]) and specificity (.89; [.86, .91]) against polysomnography. Similar sensitivity (.96; [.94, .99]) and specificity (.84; [.78 .91]) were found comparing the Actigpatch to the Motionlogger. Comparing the devices’ validity with PSG, sensitivity was not statistically different between the Actigpatch and Motionlogger (b=.0041, t=0.56; p=.58); however, the Motionlogger demonstrated higher specificity (.95; [.92, .97]) compared to the Actigpatch (b=0.065, t=4.69; p<.001). To that end, TST estimates were longer (p=.016) for the Actigpatch (449min; [428, 471] relative to the Motionlogger (438min; [416, 459]).

Conclusion

These data indicate that the adhesive “Actigpatch” is as sensitive to detect polysomnographic-confirmed sleep as a common research-grade actigraph. The Actigpatch may be less capable of detecting wake episodes. Unlike traditional actigraphs, the Actigpatch can be worn continuously for 3 weeks without risk of water or impact damage. Participants are not responsible for remembering to wear the device. Field studies, or studies in populations struggling with adherence (e.g., children) may benefit from wearable monitors such as the Actigpatch.

Support (if any)

R01AA025593, Circadian Positioning Systems

100 Effect of 3 consecutive nights of alcohol on sleep variables: Preliminary report

Introduction

The effects of a moderate dose of alcohol one hour before bedtime on sleep have not often been studied nor is the effect across nights well known. We therefore sought to test whether such effects as sleep disruption, increased early-night slow wave sleep (SWS), and reduced early-night REM sleep would be sustained across nights.

Methods

Twenty-five healthy participants (13 male; ages 22–69 yr, mean = 35) reporting moderate drinking kept a fixed sleep schedule (8–9 h TIB, confirmed by actigraphy) for about one week before two 3-night sleep studies in the lab separated by ≥ 3 days. Participants drank either mixer alone or a beverage containing alcohol targeting a breath alcohol content (BrAC) of 0.08% in a counter-balanced order over 45 min ending 1 hr before lights out. Sleep was scored using Rechtschaffen & Kales (1968) rules in 30-sec epochs. Mixed-effects models examined beverage type, study night, and the interaction of beverage and night for 13 variables: sleep efficiency, sleep latency, REM latency, and full-night percent of Stage 1, Stage 2, SWS, and REM sleep; and percent of SWS and REM sleep by thirds of night.

Results

A significant effect of Night was seen for sleep efficiency (F(2,120)=3.79; p=.025) and sleep latency (F(2,120)=5.19;p=.007), both lower on N1, as well as for REM latency, longer on N1 (F(2,120)=6.52;p=.002). REM latency was longer with alcohol (F(1,120)=14.16; p<.000) and no interaction was apparent. St2% was higher (F(1,120)=4.47; p=.037) and REM% lower (F(1,120)=4.41; p=.038) with alcohol, whereas overnight SWS% was unaffected; none showed an effect of night or an interaction. SWS% in the first (F(1,120)=10.51; p=.002) and second thirds (F(1,120)=8.27; p=.005) of the night was higher with alcohol and unaffected in the last third. REM% in the first third alone was higher with alcohol (F(1,120)=10.71; p=.01).

Conclusion

These findings show only modest effects of pre-sleep alcohol consumption (targeting 0.08% BrAC) on subsequent sleep in healthy drinkers, with no evidence of a cumulative impact across three nights. We aim to increase the sample size and examine effects on next-day cognitive function in subsequent analyses.

Support (if any)

R01AA025593

238 Adolescent Sleep Variability, Social Jetlag, and Mental Health during COVID-19: Findings from a Large Nationwide Study

Introduction

Adolescents are vulnerable to short, insufficient sleep stemming from a combined preference for late bedtimes and early school start times, and also circadian disruptions from frequent shifts in sleep schedules (i.e., social jetlag). These sleep disruptions are associated with poor mental health. The COVID-19 pandemic has impacted education nationwide, including changes in instructional formats and school schedules. With data from the Nationwide Education and Sleep in TEens During COVID (NESTED) study, we examined whether sleep variability and social jetlag (SJL) during the pandemic associate with mental health.

Methods

Analyses included online survey data from 4767 students (grades 6-12, 46% female, 36% non-White, 87% high school). For each weekday, participants identified if they attended school in person (IP), online-scheduled synchronous classes (O/S), online-no scheduled classes (asynchronous, O/A), or no school. Students reported bedtimes (BT) and wake times (WT) for each instructional format and for weekends/no school days. Sleep opportunity (SlpOpp) was calculated from BT and WT. Weekday night-to-night SlpOpp variability was calculated with mean square successive differences. SJL was calculated as the difference between the average sleep midpoint on free days (O/A, no school, weekends) versus scheduled days (IP, O/S). Participants also completed the PROMIS Pediatric Anxiety and Depressive Symptoms Short Form. Data were analyzed with hierarchical linear regressions controlling for average SlpOpp, gender, and school-level (middle vs high school).

Results

Mean reported symptoms of anxiety (60.0 ±9.1; 14%≧70) and depression (63.4 ±10.2; 22%≧70) fell in the at-risk range. Shorter average SlpOpp (mean=8.3±1.2hrs) was correlated with higher anxiety (r=-.10) and depression (r=-.11; p’s<.001) T-scores. Greater SlpOpp variability was associated with higher anxiety (B=.71 [95%CI=.41-1.01, p<.001) and depression (B=.67 [.33-1.00], p<.001) T-scores. Greater SJL (mean=1.8±1.2hrs; 94% showed a delay in midpoint) was associated with higher anxiety (B=.36 [.12-.60], p<.001) and depression (B=.77 [.50-1.03], p<.001) T-scores.

Conclusion

In the context of system-wide education changes during COVID-19, students on average reported at-risk levels of anxiety and depression symptoms which were associated with greater variability in sleep opportunity across school days and greater social jetlag. Our findings suggest educators and policymakers should consider these sleep-mental health associations when developing instructional formats and school schedules during and post-pandemic.

Support (if any)

T32MH019927(P.W.)

237 Sleep disturbances, online instruction, and learning during COVID-19: evidence from 4148 adolescents in the NESTED study

Introduction

COVID-19 fundamentally altered education in the United States. A variety of in-person, hybrid, and online instruction formats took hold in Fall 2020 as schools reopened. The Nationwide Education and School in TEens During COVID (NESTED) study assessed how these changes impacted sleep. Here we examined how instruction format was associated with sleep disruption and learning outcomes.

Methods

Data from 4148 grade 6-12 students were included in the current analyses (61% non-male; 34% non-white; 13% middle-school). Each student’s instructional format was categorized as: (i) in-person; (ii) hybrid [≥1 day/week in-person]; (iii) online/synchronous (scheduled classes); (iv) online/asynchronous (unscheduled classes); (v) online-mixed; or (vi) no-school. Sleep disturbances (i.e., difficulty falling/staying asleep) were measured with validated PROMIS t-scores. A bootstrapped structural equation model examined how instructional format and sleep disturbances predict school/learning success (SLS), a latent variable loading onto 3 outcomes: (i) school engagement (ii) likert-rated school stress; and (iii) cognitive function (PROMIS t-scores). The model covaried for gender, race-ethnicity, and school-level

Results

Our model fit well (RMSEA=.041). Examining total effects (direct + indirect), online and hybrid instruction were associated with lower SLS (b’s:-.06 to -.26; p’s<.01). The three online groups had the strongest effects (synchronous: b=-.15; 95%CI: [-.20, -.11]; asynchronous: b=-.17; [-.23, -.11]; mixed: b=-.14; [-.19, -.098]; p’s<.001). Sleep disturbance was also negatively associated with SLS (b=-.02; [-.02, -.02], p<.001). Monte-carlo simulations confirmed sleep disturbance mediated online instruction’s influence on SLS. The strongest effect was found for asynchronous instruction, with sleep disturbance mediating 24% of its effect (b = -.042; [-0.065, -.019]; p<.001). This sleep-mediated influence of asynchronous instruction propagated down to each SLS measure (p’s<.001), including a near 3-point difference on PROMIS cognitive scores (b = -2.86; [-3.73, -2.00]).

Conclusion

These analyses from the NESTED study indicate that sleep disruption may be one mechanism through which online instruction impacted learning during the pandemic. Sleep disturbances were unexpectedly influential for unscheduled instruction (i.e., asynchronous). Future analyses will examine specific sleep parameters (e.g., timing) and whether sleep’s influence differs in teens who self-report learning/behavior problems (e.g., ADHD). These nationwide data further underscore the importance of considering sleep as educators and policy makers determine school schedules.

Support (if any):

050 Evening Alcohol Consumption and Slow Wave Sleep: Impact on Morning Hippocampus-Dependent Learning across Three Nights

Introduction

Numerous studies interrogated the relationship between alcohol and a single night of sleep. Yet, many adults engage in cumulative days of drinking. Previous studies show alcohol on a single night increases slow wave sleep in the first third of night. Similarly, sleep has been associated with the success of daytime learning. Our goal was to investigate across three consecutive nights how evening alcohol use and nighttime sleep are associated with morning learning.

Methods

23 adults (11F, mean age 33.5±12 years) completed six nights of PSG monitored sleep. Participants consumed alcohol with a target 0.08 breath alcohol concentration (BrAC) and no alcohol on three consecutive nights in counterbalanced order. Percent of slow wave sleep (SWS%) in the first third of the night was derived. Learning was assessed each morning with distinct stimuli on the Mnemonic Similarity Task (MST). The MST score derived was the Lure Discrimination Index (LDI), defined as the proportion of similar images correctly identified minus the proportion of old images incorrectly identified.

Results

SWS% during the first third of the night was greater for alcohol nights compared to non-alcohol nights (F(1, 110)=10.891, p=0.01). However, there was no evidence that either night number or the interaction of drink content and night number affected %SWS in the first third of night (all p’s &gt; 0.05). There was a modest decrease in LDI on mornings following alcohol consumption; however, this effect was not significant. In a separate linear mixed-effect model we found no evidence for an effect of night number, drink content, or their interaction on MST LDI scores (all p’s &gt; 0.05).

Conclusion

Our results indicate that slow wave sleep in the first third of the night is sensitive to evening alcohol consumption. Despite prior literature associating slow wave sleep with next-day learning, we observed no effect of alcohol or night number on morning learning. It is possible that the small sample size contributed to our results. There is little prior research on the cumulative effects of alcohol on sleep and learning; our study adds to this area of research despite the negative findings.

Support (if any)

R01AA025593

232 COVID stress and sleep disturbance among a racially/ethnically diverse sample of adolescents: Analysis from the NESTED study

Introduction

Using data from the Nationwide Education and Sleep in TEens During COVID (NESTED) study (N=6,578), we investigated if race/ethnicity (64.6% were White and 35.4% identified as a racial/ethnic minority, mixed, or “other”) and community social vulnerability affected the association between COVID stress and sleep disturbance.

Methods

Data on sociodemographic factors (age, race, sex, grade, zip code [for neighborhood social vulnerability index, SVI]), COVID-related stress, depression, anxiety, instructional format (online, in-person, or hybrid), and sleep disturbance (PROMIS Pediatric Sleep Disturbance) were captured through an online survey. Descriptive and inferential analyses (Hierarchical Binary Logistic Regression (HBLR), SPSS v. 25) in 4171 adolescents examined associations between sleep disturbance and COVID-related stress, adjusting for race, sex, SVI, grade level, learning format, household density, and mental health factors.

Results

Sleep disturbance was prevalent among adolescents (89% above average, T-score &gt;50); about two-thirds (64.4%) reported greater stress due to the pandemic. Compared to White (88.5%) adolescents, sleep disturbance was more common in Black (91.2%), Hispanic (90.5%), American Indian/Alaska native (95.1%), and Mixed (92.3%) and less common in Asian (83.9%) adolescents. Chi-square analysis indicated that both race/ethnicity (□2 = 14.96, p&lt;.05) and SVI (□2 = 8.34, p&lt;.05) had an effect on sleep disturbance. HBLR analysis indicated that compared to pre-pandemic, adolescents reporting “little stress” (OR=.70, 95% CI= .49-.99, p=.04) or “the “same amount of stress” (OR=.64, 95% CI= .47-.89, p=.007) had lower odds of sleep disturbance. Higher depression (OR=1.06, 95% CI=1.04-1.07, p&lt;.001) and anxiety (OR=1.05, 95% CI=1.04-1.07, p&lt;.001) symptoms increased odds of sleep disturbance, while male gender lowered odds of sleep disturbance (OR=.11, 95% CI=.015-.86, p&lt;.05). Overall, race/ethnicity (p=.44) and SVI (p=.45) did not independently predict sleep disturbance. Race/ethnicity stratified analyses indicated that for Black and Hispanic adolescents, being in grades 11/ 12 and depression predicted sleep disturbance; and for Asian adolescents SVI and anxiety predicted sleep disturbance.

Conclusion

COVID-related stress and symptoms of depression and anxiety are associated with sleep disturbance. We observed differences in sleep disturbance across racial/ethnic groups and neighborhood social vulnerability strata, for specific racial/ethnic groups.

Support (if any)

AS was supported by funding from the NIH (K01HL135452, R01HL152453)

312 Associations Between Sleep Regularity and Body Mass Index: Findings from a Prospective Study of First-Year College Students

Introduction

College students often experience irregular sleep timing, short sleep duration, and weight gain. Using data from a large, prospective study on sleep in first-year college students, we examined whether students’ sleep regularity index (SRI; Phillips et al., 2017) was associated with body mass index (BMI) and BMI change (∆BMI) during the first nine weeks of their college semester.

Methods

Analyses included data from 583 students (mean age = 18.7± 0.5 years; 59% Female; 48% non-White) who had their height and weight assessed at the start of classes (T1) and end (T2) of nine weeks. ∆BMI was calculated as the difference between T2 and T1, with a positive value indicating an increase in BMI. Throughout the semester, participants completed on-line daily sleep diaries that included bedtime, wake-time, sleep onset latency, and wake after sleep onset for the previous major sleep episode and daytime naps. Based on this data, total sleep time (TST) was calculated as time spent asleep between bedtime and wake-time, and SRI was calculated by comparing participants’ sleep/wake states across adjacent 24-hour periods. Average SRI reflects participants’ sleep regularity (0 (random) to 100 (perfect regularity)) across the study. Data were analyzed with hierarchical linear regressions that controlled for sex and average TST.

Results

Average SRI was 74.1±8.7 (range 25.7–91.6). Average BMI at T1 was 22.0±3.5; 6% of participants were underweight (BMI less than 18.5), 6% overweight (≥25 and &lt;30) and 3% obese (≥30). Greater BMI at T1 was correlated with less ∆BMI by T2 (r=-.16, p&lt;.001). On average, participants gained 1.8±2.4kg (range: -7.2–11.4); 6% of participants lost ≥2kg, 39% gained 2-5kg, 8% gained more than 5kg. Average TST was not significantly correlated with BMI or ∆BMI. Lower SRI was associated with greater BMI at T1 (B= -.06 [95% CI: -.09– -.02], p=.001) but less ∆BMI (B= .01 [.002–.018], p=.018).

Conclusion

We found that lower sleep-wake regularity associated with greater baseline BMI but less BMI increase during the initial transition to college. Given that the majority of our participants were normal weight young adults, our findings may indicate that sleep regularity associates with healthy growth in this population.

Support (if any)

R01MH079179, T32MH019927(P.W.)

675 COVID-19 Instruction Style (In-Person, Virtual, Hybrid), School Start Times, and Sleep in a Large Nationwide Sample of Adolescents

Introduction

The COVID-19 pandemic significantly disrupted how and when adolescents attended school. This analysis used data from the Nationwide Education and Sleep in TEens During COVID (NESTED) study to examine the association of instructional format (in-person, virtual, hybrid), school start times, and sleep in a large diverse sample of adolescents from across the U.S.

Methods

In October/November 2020, 5346 nationally representative students (grades 6–12, 49.8% female, 30.6% non-White) completed online surveys. For each weekday, participants identified if they attended school in person (IP), online-scheduled synchronous classes (O/S), online-no scheduled classes (asynchronous, O/A), or no school. Students reported school start times for IP or O/S days, and bedtimes (BT) and wake times (WT) for each applicable school type and weekends/no school days (WE). Sleep opportunity (SlpOpp, total sleep time proxy) was calculated from BT and WT. Night-to-night sleep variability was calculated with mean square successive differences.

Results

Significant differences for teens’ sleep across instructional formats were found for all three sleep variables. With scheduled instructional formats (IP and O/S), students reported earlier BT (IP=10:54pm, O/S=11:24pm, O/A=11:36pm, WE=12:30am), earlier WT (IP=6:18am, O/S=7:36am, O/A=8:48am, WE=9:36am), and shorter SlpOpp (IP=7.4h, O/S=8.2h, O/A=9.2h, WE=9.2h). Small differences in BT, but large differences in WT were found, based on school start times, with significantly later wake times associated with later start times. Students also reported later WT on O/S days vs. IP days, even with the same start times. Overall, more students reported obtaining sufficient SlpOpp (>8h) for O/S vs. IP format (IP=40.0%, O/S=58.8%); when school started at/after 8:30am, sufficient SlpOpp was even more common (IP=52.7%, O/S=72.7%). Greater night-to-night variability was found for WT and SlpOpp for students with hybrid schedules with >1 day IP and >1 day online vs virtual schedules (O/S and O/A only), with no differences in BT variability reported between groups.

Conclusion

This large study of diverse adolescents from across the U.S. found scheduled school start times were associated with early wake times and shorter sleep opportunity, with greatest variability for hybrid instruction. Study results may be useful for educators and policy makers who are considering what education will look like post-pandemic.

Support (if any):

096 The Association Between Sleep Regularity Index and Self-Reported Behavioral and Emotional Symptoms in Adolescents

Introduction

Among adolescents, sleep health has been associated with emotional and mood regulation, cognitive functioning, and behavior. Few studies, however, have examined the Sleep Regularity Index (SRI, Phillips et. al, 2017) and its associations with mental health and well-being in this age group. For this study, we examined whether SRI in 15-16-year-old adolescents would predict internalizing and externalizing symptoms as measured by Youth Self-Report (YSR) scores two years later. We hypothesized that a higher baseline sleep regularity would predict lower internalizing and externalizing YSR scores at the 2-year follow-up.

Methods

The sample included 32 adolescents (14 male) ages 15-16yr (mean = 15.6) at baseline and 2 years later (mean age = 17.7). Actigraphy data and YSR scores were collected at baseline, and YSR was examined at follow-up. Participant’s SRIs were calculated using 24-hour actigraphy data scored for sleep and wake. YSR T-scores of 60 or above indicate borderline clinical internalizing (n = 2) and externalizing (n = 4) symptoms at follow-up. We used linear regression modeling to determine whether baseline SRI predicted YSR scores 2 years later. Covariates included sleep start time, sleep duration, sex, and baseline YSR scores.

Results

At baseline, average SRI and YSR scores were not significantly correlated (internalizing: r = 0.10; externalizing: r = 0.24, p’s &gt; 0.1). SRI score at baseline (mean = 80.5 ± 7.4) significantly predicted YSR internalizing scores (mean = 42 ± 9) at the 2-year follow up (t(26) = 2.57, p = 0.016) but not externalizing scores (mean = 44.8 ± 10.3, t(26) = .78, p = 0.44).

Conclusion

We observed that sleep regularity was associated with internalizing symptoms two years later; however, the association was not in the expected direction: higher SRI was correlated with increased YSR internalizing scores at the 2-year follow-up. As most participants were in a healthy range for YSR scores at both assessments, a possible explanation for this finding is that those with higher SRIs have greater self-awareness in assessing their internal feelings. Future work will examine SRI values and YSR in this sample across 6 assessments acquired at 6-month intervals.

Support (if any)

AA13252 (NIH)

Effect of experimental change in children's sleep duration on television viewing and physical activity

BACKGROUND

Paediatric observational studies demonstrate associations between sleep, television viewing and potential changes in daytime activity levels.

OBJECTIVE(S)

To determine whether experimental changes in sleep lead to changes in children's sedentary and physical activities.

METHODS

Using a within-subject counterbalanced design, 37 children 8-11 years old completed a 3-week study. Children slept their typical amount during a baseline week and were then randomized to increase or decrease mean time in bed by 1.5 h/night for 1 week; the alternate schedule was completed the final week. Children wore actigraphs on their non-dominant wrist and completed 3-d physical activity recalls each week.

RESULTS

Children reported watching more television (p < 0.001) and demonstrated lower daytime actigraph-measured activity counts per epoch (p = 0.03) when sleep was decreased (compared with increased). However, total actigraph-measured activity counts accrued throughout the entire waking period were higher when sleep was decreased (and children were awake for longer) than when it was increased (p < 0.001).

CONCLUSION(S)

Short sleep during childhood may lead to increased television viewing and decreased mean activity levels. Although additional time awake may help to counteract negative effects of short sleep, increases in reported sedentary activities could contribute to weight gain over time.

A Model-Based Approach to Optimizing Ultradian Forced Desynchrony Protocols for Human Circadian Research

The human circadian system regulates internal 24-h rhythmicity and plays an important role in many aspects of human health and behavior. To investigate properties of the human circadian pacemaker such as intrinsic period and light sensitivity, experimental researchers have developed forced desynchrony (FD) protocols in which manipulations of the light-dark (LD) cycle are used to desynchronize the intrinsic circadian rhythm from the rest-activity cycle. FD protocols have typically been based on exposure to long LD cycles, but recently, ultradian FD protocols with short LD cycles have been proposed as a new methodology for assessing intrinsic circadian period. However, the effects of ultradian FD protocol design, including light intensity or study duration, on estimates of intrinsic circadian period have not, to our knowledge, been systematically studied. To address this gap, we applied a light-sensitive, dynamic mathematical model of the human circadian pacemaker to simulate ultradian FD protocols and analyze the effects of protocol design on estimates of intrinsic circadian period. We found that optimal estimates were obtained using protocols with low light intensities, at least 10 d of exposure to ultradian cycling, and a 7-h LD cycle duration that facilitated uniform light exposure across all circadian phases. Our results establish a theoretical framework for ultradian FD protocols that can be used to provide insights into data obtained under existing protocols and to optimize protocols for future experiments.

Relationship between sleepiness and general health status

One commonly used instrument for evaluating general health and functional status is the medical outcomes survey short form 36 (MOS). Scores obtained from this instrument are known to vary with chronic diseases and depression. However, the degree to which these health dimensions may be influenced by sleep quality or sleepiness is not well understood. A cross-sectional study was performed on the association between general health status, as determined by the MOS, with sleepiness, assessed using a standardized questionnaire [the Epworth sleepiness scale (ESS)] and the multiple sleep latency test (MSLT). One hundred twenty-nine subjects (68 women), aged 25-65 years, without severe chronic medical or psychiatric illnesses, underwent an overnight sleep study, followed by an MSLT (consisting of a series of four attempts at napping at 2-hour intervals), and completed the MOS and the ESS. The mean MSLT score was 11 +/- 2 minutes, (range 2-20) and the mean ESS score was 10 +/- 5 (range 0-24). Scores for the MOS dimensions "general health perceptions", "energy/fatigue", and "role limitations due to emotional problems" were correlated significantly with ESS scores (r = -0.30, -0.41, and -0.30, respectively; p values were all < 0.001). The MSLT was also significantly correlated with "energy/fatigue" (r = -0.19; p < 0.05). After considering the effects of chronic illness and/or body mass index in a multiple hierarchical regression analysis, sleepiness, as assessed by the ESS score, explained 8% of the variance in general health perceptions, 17% of the variance in energy/fatigue, 6% of the variance in the summary measure of well-being, and 3% of the variance in the summary measure of functional status. The variation of MOS scores with sleepiness, unrelated to age or chronic disease, suggests that measures of general health status may be broadly influenced by sleepiness and sleep quality. These data suggest that 1) sleepiness has an important impact on general health and functional status, specifically influencing self-perceptions regarding energy/fatigue; 2) a more specific assessment of sleepiness in general health evaluations may help explain some of the observed variability in these measures across subjects; and 3) general health measures may be useful in the evaluations of patients with sleep disorders.

Preliminary communication: intellectual deficit and sleep-related respiratory disturbance in the elderly

Polysomnography and neuropsychological tests administered to 41 nondemented male subjects (mean age, 69.5) indicated that impaired performance was associated with sleep-related respiratory disturbance. Such deficits could reflect deficits in vigilance or cortical insult resulting from nightly hypoxemia. Whether the degree of impairment observed here predicts more severe dementia will await longitudinal studies.

Longitudinal development of sleep-related respiratory disturbance in adult humans

A number of cross-sectional studies have found that sleep-related respiratory disturbance ( SRRD ) is common in elderly adults. This preliminary study reports on samples of middle-aged and elderly humans in good health whose polysomnographically recorded sleep was followed over time. Participants were selected for low levels of SRRD on Time 1 recording. Results indicated an increase in SRRD over the study period. This increase ws not limited to rapid eye movement or nonrapid eye movement sleep; tobacco usage, medication status, and weight gain could not account for the change, although in several cases, modest weight gain was associated with increased respiratory disturbance. These data imply that future studies of incidence of SRRD may reveal changes within a 10-year period, although the pathological consequences of these age-related changes, if any at all, may take a longer period to develop.

Issues in the diagnosis and treatment of insomnia

Most people attribute a restorative function to sleep. This is because experimental or clinical sleep disturbance is usually followed by annoying symptoms of fatigue and sleepiness the following day. Can these daytime changes be documented objectively? In the past several years, the Multiple Sleep Latency Test (MSLT) has been developed and validated as an objective quantitative measure of sleepiness. Multiple assessments of sleep latency yield a profile of sleepiness across the day. This profile changes in the predicted direction with acute total and partial sleep deprivation, chronic sleep deprivation, sleep satiation, and in comparisons between hypersomnia patients and controls. Sleep and wakefulness are complementary phases in the daily cycle of human existence. Adequacy of sleep and energetic wakefulness next day are interacting phases in this cycle. Insomnia can be seen as a perception of disturbed sleep with daytime consequences, but is essentially also a symptom. This paper reviews a number of issues in the diagnosis and treatment of insomnia. The dimensions, daytime consequences and longitudinal aspects of insomnia are considered. Most investigations to date have been geared towards the problem of chronic insomnia and yet we are all likely to suffer from transient insomnia at some point. Psychiatric and psychophysiological disorders have been shown to be the most frequent causes of disorders of initiating and maintaining sleep. Moreover, there is an apparent disparity between subjective and objective sleep parameters with, for example, objectively disturbed sleep in noncomplaining subjects. The criteria of hypnotic efficacy and the effects of triazolam and flurazepam on sleep and daytime alertness have been investigated in normals, chronic insomniacs and the elderly. In general, chronic insomniacs showed all degrees of daytime alertness regardless of nocturnal sleep parameters. About one-third could be classified as fully alert all day long in spite of their complaints. The effect of flurazepam and triazolam on sleep (improvement) was essentially the same. Daytime effects were most closely related to half-life. The long-acting benzodiazepine, flurazepam, impaired daytime alertness although nocturnal sleep was improved. Triazolam improved not only nighttime sleep but also daytime alertness.

Evaluation of short-term and long-term treatment of the narcolepsy syndrome with clomipramine hydrochloride

Clinical examinations, questionnaires, and 24- or 36-hour polygraphic recordings were performed on 21 adult patients with the narcolepsy syndrome to investigate the short- and long-term effects of clomipramine HCL. Cataplexy was improved by the medication, but tolerance was observed 4 1/2 months of treatment. Clomipramine HCL induced significant changes in the sleep EEG, chin EMG, and EOG. In two patients, clomipramine HCL caused a nocturnal myoclonia that produced insomnia. Sexual side effects were seen with clomipramine HCL, particularly in males. A combination of clomipramine HCL and L-Dopa apparently prevented this difficulty in one patient. A rebound of cataplexy was seen during the 15 days following withdrawal of the drug. Methysergide maleate was found to be ineffective on cataplexy in four patients.