1. Introduction

This report discusses Sleep among MAPme participants at baseline (Wave 1).Sleep duration and sleep quality are two main domains interested in this study. For sleep duration, time to sleep, time to wake up and hours of sleep slept in both weekdays and weekends were reported by subjects. For sleep quality, participants were asked “During the past month, how would you rate your sleep quality overall? -1. Very Good -2. Good -3. Poor -4. Very Poor -NA. Perfer Not to Answer”.

This report is composed of descriptive tables and interactive plots concerning comparison of sleep duration between weekdays and weekends, different campus and gender, and sleep quality between different campus and gender.

2. Data Qulity Checking

Out of 303 subjects who compeleted the baseline assessment, 22 subjects were with plausible reported time to sleep/wake up, for instance, record 21 reported that he went to sleep at 11:00am on a typical weekdays, which could be explained by the subject chose pm rather than am accidently(Table 1).
Table 1.Plausible Self-reported Bedtime and Wake Time, n=22
ID Time to Bed (Weekdays) Time to Bed (Weekends) Time to Wakeup (Weekdays) Time to Wakeup (Weekends)
21 22.5 11 8 9.5
53 1 2 1 11
65 0 0.5 0 0.5
82 22.5 10.5 7.5 8
101 12.5 12.5 7 8.5
114 23 11 7.5 7.5
120 22.5 11 7.5 8
156 1 2 10.5 2.5
159 23 0 19.5 9
166 3 3.5 9 1
181 11 0.5 9 9.5
191 0.5 2.5 8.5 23.5
197 0 1 7 1
214 2 3 2 3
230 23.5 12.5 8 8.5
231 0 1.5 7.5 22
239 2 2 2 2
355 12 4 8 2
357 23 11.5 7.5 8
414 12.5 2 8 9
415 2 4.5 9 23
457 3 6 11 16
Note: Self-reported bedtime and wake time were record using 24-hour clock and plausible data was marked red in the table.

3. Sleep Duration

This portion mainly illustrates sleep duration calculated based on self-reported bedtime and wake time.The formula is as follows: \[ sleep duration= \begin{cases} 24 - bedtime + wake time,& \text{if } bedtime\leq 24\\ bedtime - wake time, & \text{otherwise} \end{cases} \]

3.1 Camparison of Sleep Duration in Weekdays and Weekends

The average sleep duration in weekdays is 7.83 hours, which is shoter than the sleep duration in weekends, 8.25 hours. The difference is statistically significant (95%CI: -0.62, -0.20, P-value < 0.0001, Figure 1 & Table 2).

Table 2.Plausible Self-reported Bedtime and Wake Time
Category n Mean SD p.value Lower CI Upper CI
Weekdays 281 7.83 1.21 2e-04 -0.62 -0.2
Weekends 281 8.24 1.36
Note: Two sample student t-test was applied to compare the sleep duration in two categories

3.1.2 Sleep Duration in Weekdays

The distribution of sleep duration in weekdays is approximately normally distributed.Half of the subjects sleep less than 8 hours during the weekdays. Compared to male subjects, felmale subjects are more likely to report shoter sleep (7.81 h V.S. 7.88 h).The average sleep duration among subjetcs in Emory campus is higher than that among subjects in Oxford campus(7.88 h V.S. 7.44 h), while the range of sleep duration among subjects in Oxford campus is wider . Student t-test was applied to exam the difference between genders and campus, indicating not statistically significant.(Figure 2-4)

3.1.3 Sleep Duration in Weekends

The distribution of sleep duration in weekends is normally distributed.Half of the subjects sleep less than 8.5 hours during the weekend.Similar to the average sleep duration in weekdays, the average sleep duration in weekends among subjetcs in Emory campus is longer than that among subjects in Oxford campus(8.28 h V.S. 8.17 h). However, the gender difference in sleep duration varied, female reporting sleep more than male in weekends(8.19 h V.S. 8.34 h). (Figure 5-7)

4. Sleep Quality

This portion discussed descriptive result of sleep quality as well as comparison of gender and campus difference between sleep quality categories. Out of 281 subjects, 182(64.77%) reported having good sleep quality, only 13(4.63%) reported having very poor sleep quality , 54(19.33%) and 32 (11.39%) reported having poor and very good sleep quality, separately.There is no statistically significant gender and campus difference within categories of sleep quality(Both P values are greater than 0.05).Subjects who studied in Emory Unversity Atlanta Campus are more likely to report “good” or “very good” sleep quality than those studied in Emory Univerity Oxford Campus, which is corresponding to shorter sleep duration among subjects studied in Emory Univeristy Oxford Campus.(Table 3-5 & Figure 8-9)
Table 3.Frequency Table of Sleep Quality
Sleep Quality n prop
Very Good 32 0.1139
Good 182 0.6477
Poor 54 0.1922
Very Poor 13 0.0463
Table 4.Frequency Table of Sleep Quality and Gender
Female Male
Very Good 21 11
Good 130 51
Poor 38 16
Very Poor 10 3
Note: Fish’s exact test was applied to compare the gender difference of sleep quality. P-value =0.8820.
Table 5.Frequency Table of Sleep Quality and Campus
ECAS Oxford
Very Good 22 10
Good 120 62
Poor 32 22
Very Poor 9 4
Note: Fish’s exact test was applied to compare the campus difference of sleep quality. P value = 0.7713.

5. Report Sleep Duration and Calculated Sleep Duration

This portion is mainly about the relationship between self-reported sleep duration and calculated sleep duration in weekdays and weekedns, as well as whether the relationship varied within different gender and sleep quality groups. In general, the correlation between self-report sleep duration and calculated sleep duration are positive and statistically significant in female group despite sleep quality, while the correlation is not statistically significant in some male groups: correalation between self-reported sleep duration and calulated sleep duration on weekdays is not statistically significant in male groups with very good and poor sleep quality, besides male with very poor sleep quality have less consistency between self-reported sleep duration and calculated sleep duration on weekends.