📄 medicine

As Digital Behavior and Lifestyle Factors in Relation to Sleep Quality and Mental Fatigue: A Secondary Cross-Sectional Analysis of an Open Dataset

This secondary cross-sectional analysis of 15,000 adults reveals that while pre-sleep smartphone use and stress levels are independently associated with sleep quality and mental fatigue, the extremely strong correlation between stress and mental fatigue suggests significant construct overlap, and the study's cross-sectional design limits conclusions to associations rather than causation.

Original authors: Kannadasan Karuppaiah, Raji Kaliyaperumal, Ayesha Siddiqha Mukthar, Vinoth Raman

Published 2026-07-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Kannadasan Karuppaiah, Raji Kaliyaperumal, Ayesha Siddiqha Mukthar, Vinoth Raman

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine your brain is a high-performance video game console. You've been playing hard all day, and now it's time to shut it down, clear the cache, and recharge the batteries for tomorrow. But what if, instead of powering down, you keep the screen glowing right up until the moment you try to sleep? That's the story this study tells, based on a massive look at 15,000 people from all walks of life—from doctors and software engineers to students and teachers.

The researchers didn't invent new data; they acted like digital detectives, digging through a giant, public treasure chest of 15,000 records to see what habits were linked to how well people slept and how tired their minds felt. Here is what they found, and just as importantly, what they didn't find.

The Big Two Culprits: Stress and the Midnight Glow

The study suggests that two things are the biggest troublemakers for your sleep and mental energy: stress and pre-sleep phone use.

Think of your sleep quality like a dimmer switch on a lamp. The study found that the more stress you have, the darker the lamp gets (a very strong link, where the numbers show a relationship of -0.857). It's like stress is a heavy blanket smothering your ability to rest.

Then there's the phone. The researchers found that using a smartphone in the 60 minutes before bed is like trying to sleep while someone is shining a flashlight in your eyes. The data shows a weak but clear negative link: the more you scroll right before bed, the worse your sleep quality tends to be. It's not a giant, world-ending correlation, but it's a consistent nudge in the wrong direction.

The "Tired Brain" and the "Stressed Brain" are Best Friends

Here is a weird twist: the study suggests that "mental fatigue" (feeling mentally drained) and "stress" are so closely linked in this dataset that they almost look like the same thing. The numbers show they are nearly identical twins, with a connection score of 0.949.

Imagine stress and mental fatigue as two dancers moving in perfect, unison. Because they move so similarly in this data, the researchers couldn't separate them to see which one was leading the dance. They suggest this might be because people report them in the same way, or because they genuinely feed off each other so much that they blur together. The paper is careful to say this doesn't prove they are the same thing, but in this specific group of 15,000 people, they acted like it.

The "Myth Busters": What Didn't Matter?

This is where the study gets really interesting because it rules out some things we often think are the main villains.

Boys vs. Girls: You might think boys and girls sleep differently or get stressed differently. But in this massive group of 15,000, the study found no meaningful difference. The numbers for sleep, stress, and fatigue were practically identical for men and women. The difference was so tiny (less than 0.03 on a scale where 0.20 is considered "small") that it's like trying to tell the difference between two grains of sand. The paper suggests that for this group, being male or female didn't predict how tired you'd be.
Age: Did older people sleep worse? Did younger people stress more? The study found no link between age and these outcomes. The average age of the group was about 38 years old, and the data suggests that in this crowd, being 20 or 50 didn't change your sleep score.
Exercise: This one is surprising. We often hear that exercise helps you sleep. But in this dataset, the amount of physical activity people reported (about 59 minutes a day) had no connection to how well they slept. The paper suggests this might be because the way people reported their exercise was too vague (like just saying "I moved a bit") to catch the real effect, or maybe the "digital glow" was just so strong it drowned out the benefits of exercise.

How Sure Are They?

The authors are very careful not to shout "We found the cure!" Instead, they say these findings suggest a pattern. Because the study looked at a snapshot in time (a cross-section) rather than following people over years, they can't say for sure that the phone caused the bad sleep, or that stress caused the fatigue. It's like seeing a photo of a messy room and a pile of clothes; you know they are together, but you don't know if the clothes made the mess or if the mess made you drop the clothes.

Also, the data came from people filling out surveys themselves. The researchers admit they don't know exactly how the questions were asked or if the tools used to measure "stress" were perfect. This means the results are a strong hint—a hypothesis generator—rather than a final, unshakeable law of physics.

The Takeaway

If you want to improve your sleep and clear your mental fog, the study suggests focusing on digital hygiene (putting the phone away before bed) and managing stress. It suggests that worrying about your gender, your age, or even how much you exercised might not be the most important lever to pull right now.

The paper ends by saying this is a starting point. It's a map that shows where the treasure might be, but we need better tools and longer journeys (long-term studies) to dig it up and prove exactly how it works. For now, the message is simple: in a world of screens and stress, those two factors seem to be the loudest voices in the room when it comes to how well we rest.

Technical Summary: Digital Behavior, Lifestyle Factors, and Sleep/Mental Health

Problem Statement
The rapid proliferation of digital technologies has fundamentally altered human behavior and lifestyle, creating a public health challenge regarding sleep structure and mental well-being. While epidemiological data indicates that over 70% of adults engage with electronic devices within an hour of bedtime, there is a lack of mechanistic and epidemiological evidence regarding the combined impact of digital behavior and lifestyle factors on sleep quality and mental fatigue, specifically through the lens of psychological stress. Previous literature has largely relied on convenience samples of students or young adults, often failing to include diverse occupational groups or sufficiently large sample sizes to detect subtle variations. Furthermore, critical moderating variables, such as physical activity, have frequently been omitted from models linking screen time to sleep outcomes.

Methodology
This study is a secondary cross-sectional analysis of a fully anonymized, open-access dataset containing 15,000 records. The sample comprises eight distinct occupational groups: graphic designers, doctors, freelancers, managers, researchers, software engineers, students, and teachers.

Data Source: The data was sourced from a public repository (Kaggle). The authors did not recruit participants; all variables were analyzed as pre-existing, self-reported indices.
Variables:
- Digital Behavior: Defined specifically as the average time (minutes) spent using a smartphone in the 60 minutes prior to sleep onset.
- Outcomes: Sleep quality, stress level, and mental fatigue. These were provided as pre-computed composite indices on a 0–10 scale. The study explicitly notes that the original measurement instruments and their psychometric properties (e.g., Cronbach's alpha) are undocumented by the data provider; thus, these are treated as self-reported indices rather than scores from validated scales.
- Covariates: Age, gender, and physical activity (self-reported minutes of moderate-to-vigorous activity per day).
Statistical Analysis: Analyses were conducted using SPSS version 31.0. Given the large sample size ( $N=15,000$ $N = 15, 000$ ), the study prioritized effect sizes (Cohen's $d$ $d$ , Cramer's $V$ $V$ , $r$ $r$ , $R^2$ $R^{2}$ ) over $p$ $p$ -values to avoid identifying trivial differences as statistically significant.
- Tests: Independent samples $t$ -tests for gender differences, Chi-square tests for occupational distribution, Pearson product-moment correlations for linear associations, and multiple linear regression to identify independent predictors.
- Collinearity: Variance Inflation Factors (VIF) were calculated. Stress and mental fatigue were found to be near-collinear ( $r = 0.949$ ) and were not entered simultaneously into the same regression model to prevent unstable coefficients.

Key Results

Demographics and Gender: The sample showed a balanced gender distribution across all eight occupational categories. There were no statistically significant differences between men and women regarding sleep quality, stress levels, or mental fatigue. All gender effect sizes were negligible (Cohen's $d \leq 0.03$ ).
Correlations:
- Sleep Quality: Showed a very strong negative correlation with stress ( $r = -0.857$ ) and mental fatigue ( $r = -0.808$ ).
- Stress and Fatigue: These two constructs were nearly perfectly positively correlated ( $r = 0.949$ ), suggesting significant construct overlap or common method variance within this dataset.
- Digital Behavior: Pre-sleep phone use showed a weak negative correlation with sleep quality ( $r = -0.335$ ) and negligible positive correlations with stress ( $r = 0.155$ ) and mental fatigue ( $r = 0.150$ ).
- Null Findings: Age and physical activity demonstrated negligible and non-significant associations with all primary outcomes ( $|r| \leq 0.013$ ).
Regression Analysis:
- A multiple linear regression model predicting sleep quality ( $R^2 = 0.78$ ) identified pre-sleep phone use ( $\beta = -0.207$ ) and stress level ( $\beta = -0.825$ ) as the only independent predictors. Age and physical activity were not significant.
- A parallel model for mental fatigue indicated that sleep quality was the most prominent predictor ( $\beta = -0.853$ ).

Key Contributions

Large-Scale Occupational Diversity: Unlike prior studies focused on students, this analysis provides a broad view across eight diverse professional groups, suggesting that the associations between digital behavior and sleep/fatigue are consistent across different occupational contexts.
Quantification of Effect Sizes: By emphasizing effect sizes over $p$ -values in a massive dataset, the study clarifies that while pre-sleep phone use is statistically associated with poorer sleep, the magnitude of this specific association is weak compared to the overwhelming influence of stress.
Identification of Construct Overlap: The study highlights the near-perfect correlation between self-reported stress and mental fatigue in this specific dataset, cautioning against treating them as entirely distinct independent variables in future modeling without validated, distinct measurement tools.
Null Results on Lifestyle Factors: The analysis explicitly reports the lack of association between physical activity and sleep outcomes in this specific self-reported dataset, attributing this likely to the crudeness of the measurement (aggregate self-report) rather than a true absence of biological effect.

Significance and Claims
The authors position this work as hypothesis-generating rather than definitive proof of causality. The study claims that:

Primary Drivers: In this specific dataset, pre-sleep phone use and psychological stress are the factors most strongly associated with sleep quality and mental fatigue.
Nature of Evidence: Due to the cross-sectional design and reliance on self-reported composite indices, the findings indicate associations, not predictors or causes. The study does not claim to prove that phone use causes poor sleep, but rather that they are linked.
Public Health Implications: The results support the need for public health strategies focused on "digital hygiene" (e.g., screen-free periods before bed) and stress reduction to address the growing burden of sleep disorders.
Limitations as Caveats: The authors explicitly state that the lack of validated instruments, the potential for common method variance, the absence of clinical exclusion criteria (e.g., diagnosed sleep disorders), and the lack of demographic context (geography, socioeconomic status) limit the generalizability of the findings. The study calls for future research using objective measures (actigraphy, ecological momentary assessment) and longitudinal designs to confirm causal mechanisms.