A Data Driven Study of Sleep and Associated Factors

Overview
In this project, we will explore the relationship between sleeping patterns (sleeping duration, and quality of sleep) and various demographic variables (age, and gender) and lifestyle factors (physical activity level, daily steps, stress level, heart rate and so on) using the Sleep Health and Lifestyle dataset obtain from Kaggle. We will be using the data visualizations and statistical interpretations to uncover the meaningful insights into how lifestyle and personal characteristics influence sleep health.

Goal
To analyze and visualize how sleep duration and sleep disorders are related to different lifestyle and demographic variables.

Objectives
- Analyze and compare the distribution of sleep duration across gender and sleep disorder status using histograms and boxplots.
- Explore sleep disorder variation by gender and BMI categories using bar plots.
- Identify correlations between age, sleep duration, physical activity, sleep quality, stress level, daily steps, and heart rate using a heatmap visualization.
- Observe trends between sleep quality and stress level/heart rate using scatter plots with regression lines.

Histogram suggests a wide range of sleeping duration. Most individuals sleep between 6.5 to 7.5 hours. The distribution is not normal. It is bi-modal and skewed.It is bi-modal and skewed. It has two noticeable peaks,one at 7.0 hours and one in between 7.5 and 8.0 hours indicating two common sleep duration patterns. Large number of individuals sleep approximately 7 hours which is a recommended sleeping hours for an adult. Few people sleep more than 8.0 hours.

Sleeping duration distribution is negatively skewed for males and positively skewed for females. Even though the median for both groups is the same, females have a higher IQR (Q3-Q1) than males, indicating more variation in sleeping duration. Sleep duration for females is higher than that of males.

Individuals without a sleeping disorder sleep between 6.5 and 8 hours, with a median of 7.5 hours, which is an ideal sleep cycle. Individuals with Sleep Apnea have longer sleep durations and variations, and the median is close to 8 hours. Insomnia is related to reduced sleep duration. Individuals with Insomnia have shorter sleep durations, and most values are concentrated around 6.5 hours, with several outliers below 6 hours.

A higher proportion of males do not have any sleep disorders compared to females. Sleep apnea is much more common among females. Nearly the same number of both genders have reported insomnia.

More males have a normal BMI than females. Very low individuals are categorized in the obese category, but more males are in this category than females. Overweight is much more common in females than in males.

Most individuals with a normal BMI have no sleeping disorder. Insomnia and sleep apnea are common in overweight individuals. Very few individuals with normal weight have reported insomnia and sleep apnea.

Age has weak negative correlation with physical activity level, daily steps and quality of sleep indicating daily activity and quality of sleep decreases with age. Stress level and heart rate are negatively correlated with sleep duration and quality of sleep, suggesting better sleep is associated with lower stress and heart rate. Quality of sleep is positively associated with daily steps means higher number of daily steps leads to better sleep. Physical activity level negatively related with stress and heart rate, suggesting active individuals experience less stress and lower heart rate.

The scatter plot between stress level and quality of sleep suggests a storng negative association. In other words, if the stress level increases quality of sleep is likely to decrease. Points are distributed above and below the fitted line (green line), indicating a negative linear relationship. The dashed red lines represent confidence intervals.

The plot suggests a moderate negative correlation between heart rate and quality of sleep. Most of the points lie close to the straight line. The red dashed line shows the confidence level. Very few points lie far from the straight line and the main cluster, indicating the presence of outliers or extreme values.

Conclusions:

• Even though median sleep duration for both females and males are same, females tend to sleep longer and show more variation in sleep duration than males
• Males are more likely to have normal BMI than females.
• Overweight is more common in females.
• Sleep disorder is more frequent in overweight individuals.
• Sleep duration is impacted by sleep disorder. For instance: Ideal sleep duration tends to no disorder and insomnia and sleep apnea lead to shorter and longer sleep duration respectively.
• Higher stress level and heart rate subject to lower sleep quality.

References:

• El Hattab, Hakim, and JJ Allaire. 2017. Revealjs: R Markdown Format for Reveal.js Presentations. https://github.com/rstudio/revealjs.
• Uom190346a. (n.d.). Sleep health and lifestyle dataset [Data set]. Kaggle. https://www.kaggle.com/datasets/uom190346a/sleep-health-and-lifestyle-dataset