“Sleeping is no mean art: for its sake one must stay awake all day.”

-Friedrich Nietzsche

Image for post
Image for post
Photo by Claudia Mañas on Unsplash

How Can Our Daily Habits Influence Sleep and Fatigue Levels?

In 1910 the average American slept 9 hours per night, whereas a Gallup poll showed that number declining to 7.9 in 1942 and to 6.8 hours per night by 2013. This is all while recent research is showing a link between poor or insufficient sleep and nearly every chronic disease that plagues us — heart disease, diabetes, Alzheimer’s and even cancer.

Despite advancements in living standards in modern society that are meant to give us back control of our time, precious little of that time or focus has gone into improving the hours or quality of sleep.

What can we do about this?

The dataset examined here was downloaded from Kaggle, and comes from a sleep study that was performed, collecting self-reported answers from 104 individuals. Each study participant answered questions regarding some of their daily habits, as well as how much they slept, and how tired they felt during the day.

The questions answered by each individual constituted the following variables in the dataset:

  • Do you get enough sleep at night? (Y/N)
  • How many hours of sleep do you usually get?
  • Do you keep your phone within arm’s reach at night? (Y/N)
  • Do you use your phone within 30 minutes of going to bed? (Y/N)
  • How tired do you feel during the day, on a scale of 1–5? (1 is not tired, 5 is most tired.)
  • Do you usually eat breakfast? (Y/N)

The specific questions that came to mind that may have some light shed upon them by analyzing this dataset are:

  1. Do people have a good sense of what constitutes “enough sleep”, or a consistent reference point for what they consider enough sleep?
  2. Do habits of phone use around bedtime and where the phone is located at night have any association with how much sleep a person gets or how tired they feel during the day?
  3. Is there any association between whether a person eats breakfast and either how many hours of sleep they get or their daytime fatigue levels?

In our dataset, the self reported number of hours of sleep had a mean of 6.65, a median of 7 and a standard deviation of 1.4. At the 25th percentile was 6 hours of sleep.

To answer the first question, I examined whether survey responders’ assessment of whether they got enough sleep was associated with the number of hours of sleep they reported. In fact, there was a sufficiently strong relationship between these answers that we can trust these peoples’ sense of whether they slept a certain number of hours, as seen in Figure 1.

Image for post
Image for post
Figure 1

Next I was interested in what we can learn about the number of hours of sleep a person is likely to get, by their phone-related behaviors.

Image for post
Image for post
Figure 2
Image for post
Image for post
Figure 3

These data did not show an association between hours of sleep and either keeping one’s phone within arm’s reach at night or using one’s phone within 30 minutes of bedtime, as seen in Figures 2 and 3. This finding may fly in the face of common sleep-hygiene wisdom. Nonetheless, it is the result of this particular study.

The strongest relationship with number of hours slept per night came not from phone use behaviors, but whether a person eats breakfast.

Since the direction of the relationship between these two variables was somewhat clear from Figure 4, I performed a one-tailed two-sample t-test. This allowed me to maximize the statistical power to detect a small effect in the direction that is apparent, i.e. that people who eat breakfast tend to report more hours of sleep, rather than fewer.

Image for post
Image for post
Figure 4

Of course we must be careful of what conclusions we draw from this notion, which may be enticing. Since this is simply an association, drawn from observational data, we cannot say that eating breakfast will lead a person to get more sleep. Nor can we say that sleeping more hours causes a person to want to eat breakfast. There is merely an association here — two phenomena that tended to be observed together, at least in this small study.

Next I wanted to evaluate whether phone use habits or breakfast eating habits had any effect on study participants’ self-reported levels of fatigue during the day.

It turned out that daytime tiredness was more sensitive to where a person’s phone is kept at night than whether they use their phone within 30 minutes of going to bed. 83% of people who kept their phones near their bedside reported daytime fatigue of level 3 or greater, whereas only 63% of people who do not keep their phones within reach at nigh reported fatigue levels as high. Figure 5 shows the variation in these proportions visually.

Image for post
Image for post
Figure 5

Daytime fatigue levels were not significantly associated with whether a person used their phone within 30 minutes of bedtime, as seen in Figure 6.

Image for post
Image for post
Figure 6

Though the relationships shown here are not strong and no causation can be inferred, these data indicate that habits of phone use may have some implications for the amount of sleep we get. It is unknown whether the individuals in this study were teens, adults or older adults, there is growing concern among sleep researchers that teens especially are at risk for health problems associated with insufficient sleep, due to the presence of technology in the bedroom.

We don’t know anything else about the individuals who participated in this study, other than that with 6 hours of sleep nightly, at least 25% of participants are under-slept according to the guidelines of the National Sleep Foundation. A larger scale study such as this one that would also include at least age, would allow us to draw even more helpful conclusions about the relationships between daily habits and sleep as discussed here.

Another limitation of this study is that all the data are observational and self-reported. Clearly this topic warrants studies of higher quality, including prospective and randomized designs, as well as collection of sleep data via wearable trackers for more objective and comparable data.

The iPython notebook created for this analysis can be found here.

I am in pursuit of better human health, quality of life, longevity and "health-span," through data driven learning.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store