1.7 Demographic and anthropometric profile

2.1 Summary of quantitative variables

2.2 Frequencies and percentages for each analytical question

Quantitative responses are grouped only for presentation; inferential tests retain original ungrouped values whenever appropriate.

2.3 Cross-tabulations

Cells show count followed by row percentage, enabling proportional comparison across groups of different sizes.

Table 2.15. Stress × sleep quality

Table 2.16. Screen time × sleep-duration category

3.1 Pearson correlations for primary hypotheses

3.2 Sleep quality by smoking status (Welch t-test)

3.3 Association between high stress and poor sleep

Test result: χ²(1) = 32.202, p < .001, Cramér V = 0.462 (moderate).

3.4 Multiple linear regression predicting sleep quality

The outcome is sleep quality (higher = better); predictors are stress, screen time, exercise days, caffeine and age. Model fit was R² = 0.764، adjusted R² = 0.755 (n = 151).

4.1 Overall sleep profile and research meaning

Mean sleep duration was 7.30 hours; mean sleep-onset latency was 52.6 minutes; mean nightly awakenings were 1.70، and mean sleep quality on the 1–5 scale was 3.68. These dimensions are complementary: duration may be adequate despite delayed initiation or repeated awakening, while short duration can coexist with a relatively favorable subjective rating.

A total of 54 participants reported under 7 hours, representing 35.8% of valid observations; 20 (13.2%) reported poor or very-poor quality. The AASM/SRS recommends that adults regularly obtain at least 7 hours [1], while an umbrella review links extreme sleep durations with a range of adverse health outcomes [12]. This remains a descriptive benchmark because individual need, illness and medication are not measured.

4.2 Hypothesis-by-hypothesis comparison with prior evidence

H1: Daily screen time is associated with sleep duration. n = 151, r = -0.421, 95% CI [-0.544, -0.280], p < .001. This is a moderate effect by magnitude. Both the direction and statistical significance align with the predominant pattern in prior literature. [2] The confidence-interval width and sample size should be weighed with the p value: a non-significant result may be imprecise, and a significant result may still be small in practical terms.

H2: Stress level is associated with sleep quality. n = 151, r = -0.811, 95% CI [-0.859, -0.748], p < .001. This is a large effect by magnitude. Both the direction and statistical significance align with the predominant pattern in prior literature. [3] The confidence-interval width and sample size should be weighed with the p value: a non-significant result may be imprecise, and a significant result may still be small in practical terms.

H3: Exercise frequency is associated with sleep quality. n = 151, r = 0.287, 95% CI [0.133, 0.427], p < .001. This is a small effect by magnitude. Both the direction and statistical significance align with the predominant pattern in prior literature. [4] The confidence-interval width and sample size should be weighed with the p value: a non-significant result may be imprecise, and a significant result may still be small in practical terms.

H4: Sleep duration is associated with self-rated productivity. n = 151, r = 0.864, 95% CI [0.817, 0.899], p < .001. This is a large effect by magnitude. Both the direction and statistical significance align with the predominant pattern in prior literature. [6] The confidence-interval width and sample size should be weighed with the p value: a non-significant result may be imprecise, and a significant result may still be small in practical terms.

H5: Sleep duration is associated with life satisfaction. n = 151, r = 0.814, 95% CI [0.752, 0.862], p < .001. This is a large effect by magnitude. Both the direction and statistical significance align with the predominant pattern in prior literature. [7] The confidence-interval width and sample size should be weighed with the p value: a non-significant result may be imprecise, and a significant result may still be small in practical terms.

4.3 Detailed discussion by exposure and outcome

Screens and sleep. A study of 122,058 adults linked daily pre-bed screen use with poorer quality, shorter duration and later bedtimes [2]. A newer meta-analysis also associated greater screen time with short sleep, insomnia symptoms and delayed bedtime, with variation across countries [10]. Plausible pathways include displacement of sleep opportunity, evening light exposure, and cognitive or emotional arousal. Brabdo measures total daily hours rather than device, content or timing, so it cannot separate those pathways.

Stress and sleep. A Saudi medical-student study found high prevalence of poor sleep and stress and a significant association between them [3]. Stress may operate through hyperarousal, rumination and difficulty disengaging from daily demands, but the relationship may be bidirectional because poor sleep can impair emotion regulation and elevate perceived stress. A meta-analysis of 65 randomized trials supports the reverse side of this loop: improving sleep produced a medium improvement in mental health and reduced stress [9]. The current single stress item is not equivalent to a validated multi-item scale.

Physical activity. A meta-analysis of 22 randomized trials found improved subjective sleep quality with exercise interventions [4]. Exercise may support mood and stress regulation, strengthen homeostatic sleep drive, and reinforce circadian timing. However, a count of exercise days does not capture intensity, duration or timing and cannot distinguish light activity from structured training; the current association is therefore simplified and not equivalent to an intervention effect.

Caffeine. A meta-analysis of 24 studies found that caffeine reduced total sleep time by about 45 minutes and sleep efficiency by 7%, while increasing sleep latency and wake after sleep onset [8]. Adenosine-receptor blockade is a plausible mechanism, but dose, timing and individual tolerance matter. Brabdo records cups rather than milligrams or timing, so exposures can differ substantially. Reverse causation is also plausible if short sleepers use caffeine to counter daytime sleepiness.

Smoking. A meta-analysis identified objective sleep-architecture disruption among smokers while the pooled subjective PSQI difference was non-significant [5]. Smoking may combine nicotine stimulation, overnight withdrawal and correlated health or social factors. The group contrast should therefore be read using Cohen’s d, its confidence interval and group sizes, recognizing that the binary measure omits dose, duration and former smoking.

Work hours, productivity and life satisfaction. Sleep-health dimensions were associated with productivity loss in a Japanese workforce [6], and a review of workers transitioning into shift work found worsening sleep and mental health [11]. Recommended and restorative sleep were also associated with greater life satisfaction in a nationally representative U.S. sample [7]. Sleep may operate through alertness, attention and mood, but direction is unresolved: work demands may restrict sleep opportunity, while poor sleep may lower perceived performance. Productivity and satisfaction are self-rated single items and may also reflect mood and response style.

4.4 Plausible mechanisms and alternative explanations

No single pathway explains every finding. Screens, caffeine and nicotine may directly affect arousal or circadian timing, whereas stress and work demands may primarily restrict sleep opportunity or increase rumination. Physical activity may be protective, but it can also mark better health or greater discretionary time. Such third variables are confounders; when unmeasured or uncontrolled, they can make an observed association stronger or weaker than the underlying relationship.

Reverse causation and feedback are also plausible. Stress may worsen sleep, while poorer sleep may produce greater next-day stress, lower productivity, more caffeine use or longer screen exposure. Because all variables are collected at one time point, neither correlations nor regression can identify where the loop began.

Common self-report measurement may inflate some associations when a participant uses a similar rating style across questions; recall error or restricted response ranges may instead attenuate them. Current findings are therefore hypothesis-generating. Confirmation requires repeated temporal measurement, wearables or validated scales, with adjustment for health, medication, shift work, age and gender.

4.5 Integrated interpretation and practical significance

Within the sample, high stress was reported by 41.7%، screen exposure of 4 or more hours by 69.5%، exercise on at least 3 days by 62.3%، 3 or more caffeine drinks by 51.0%، work of 9 or more hours by 9.9%، and smoking by 22.5%. These percentages describe exposure burden in the sample, not effect size by themselves: an exposure may be common with a weak association, or uncommon with a large but imprecise association.

With stress, screen time, exercise, caffeine and age entered together, the model explained 76.4% of sleep-quality variance in the analytical sample. This is in-sample fit, not guaranteed prediction for new participants. Predictor estimates may shift after adjustment because exposures are correlated or the sample is small; B, SE, p values and model assumptions in Part Three should therefore be reviewed together.

Practically, the findings identify areas for further assessment and education—regular sleep opportunity, stress management, less evening screen exposure, attention to caffeine timing and regular activity—but they do not prove that changing one factor will improve an individual outcome. Persistent or severe symptoms warrant professional health assessment.

4.6 Main findings

1. Daily screen time is associated with sleep duration. r = -0.421 (moderate).
2. Stress level is associated with sleep quality. r = -0.811 (large).
3. Exercise frequency is associated with sleep quality. r = 0.287 (small).
4. Sleep duration is associated with self-rated productivity. r = 0.864 (large).
5. Sleep duration is associated with life satisfaction. r = 0.814 (large).
6. High stress was associated with poor/very-poor sleep status.
7. The multivariable model explained 76.4% of sleep-quality variance in the analytical sample.

4.7 Conclusion

The study describes an interconnected network of sleep, lifestyle and wellbeing indicators in a voluntary adult sample. Current findings can refine more specific hypotheses, but self-report measurement, cross-sectional design and non-probability sampling preclude causal interpretation and unrestricted generalization.

4.8 Recommendations

• Continue targeted recruitment to improve balance across gender, age, country and employment, with an a priori power-based target sample size.
• Add screen timing, work/shift schedule, caffeine timing, health conditions and medications as potential confounders.
• Use validated multi-item instruments such as the PSQI and Perceived Stress Scale in a confirmatory study, with Arabic expert review, back-translation and pilot testing.
• Use longitudinal or intervention follow-up to test temporality and causality, and conduct sensitivity analyses for outliers, nonlinearity and interactions.
• For practice, emphasize sleep-health education, reduced pre-bed screen exposure, stress management and regular activity, with persistent symptoms referred to a health professional.

4.9 Limitations

• The voluntary online sample is vulnerable to selection and coverage bias and is not nationally representative.
• The cross-sectional design cannot determine temporal order or exclude reverse causation.
• All measures are self-reported and may be affected by recall, social desirability and differing interpretations of scales.
• Several constructs use single items; measurement error may attenuate or alter associations.
• Multiple testing increases false-positive risk; findings are exploratory and require replication and prespecified multiplicity control in confirmatory analysis.
• Assumptions of Pearson correlation, regression, Welch testing and chi-square may not hold in small samples or unusual distributions.

4.10 Comparative references

1. Watson NF, et al. (2015). Recommended Amount of Sleep for a Healthy Adult: A Joint Consensus Statement of the AASM and SRS.
2. Zhong C, et al. (2025). Electronic Screen Use and Sleep Duration and Timing in Adults. JAMA Network Open.
3. Almojali AI, et al. (2017). The prevalence and association of stress with sleep quality among medical students.
4. Xie Y, et al. (2021). Effects of Exercise on Sleep Quality and Insomnia in Adults: A Systematic Review and Meta-Analysis of RCTs.
5. Catoire S, et al. (2021). Tobacco-induced sleep disturbances: A systematic review and meta-analysis.
6. Ishibashi Y, Shimura A. (2020). Association between work productivity and sleep health: A cross-sectional study in Japan.
7. Ogbenna BT, et al. (2026). Satisfaction with life in relation to sleep health among a nationally representative sample of U.S. adults.
8. Gardiner C, et al. (2023). The effect of caffeine on subsequent sleep: A systematic review and meta-analysis.
9. Scott AJ, et al. (2021). Improving sleep quality leads to better mental health: A meta-analysis of randomized controlled trials.
10. He Z, et al. (2025). The association of screen time and the risk of sleep outcomes: A systematic review and meta-analysis.
11. Hulsegge G, et al. (2024). Sleep, mental health and physical health in new shift workers transitioning to shift work: Systematic review and meta-analysis.
12. Zhou Q, et al. (2021). Sleep duration and health outcomes: An umbrella review.

References support scientific comparison and do not imply that the Brabdo sample is equivalent to populations in those studies. Literature last reviewed: 14 August 2026.