Differential associations of the timing and amount of social media use with sleep quality and academic performance among undergraduate medical and dental students in Pakistan: a cross-sectional study.
Atif K, Ahmad A, et al. • Frontiers in public health • 2026
Sleep quality was primarily associated with when social media was used (pre-sleep timing), while examination performance was related to how much social media was used (total daily duration), demonstrating outcome specificity in a cross-sectional study of 364 Pakistani medical and dental students.
Key Findings
Results
Poor sleep quality was highly prevalent among the sampled medical and dental students.
Poor sleep quality (PSQI > 5) was present in 312 students, representing 85.7% of the sample (95% CI 81.7–88.9%).
Sleep quality was measured using the Pittsburgh Sleep Quality Index (PSQI).
The sample consisted of 364 undergraduate students enrolled in MBBS or BDS programs at ten institutions in Khyber Pakhtunkhwa, Pakistan.
Data were collected between October 2023 and April 2024 via paper-based questionnaire using convenience sampling.
Results
Duration of pre-sleep social media use was the only social media exposure independently associated with PSQI global sleep quality score.
Each one-category increase in pre-sleep social media duration was associated with a PSQI global score increase of B = 1.594 (95% CI 1.081 to 2.106, p < 0.001).
Total daily social media use duration was not independently associated with sleep quality after adjustment.
Models were adjusted for background characteristics and other social media exposures.
Nocturnal checking behavior was not associated with sleep quality outcomes.
Results
Total daily social media use duration was independently and negatively associated with academic examination scores.
Each one-category increase in total daily social media use was associated with examination scores that were 1.40 percentage points lower (B = -1.404, 95% CI -2.053 to -0.756, p < 0.001).
Academic performance was assessed using self-reported percentage marks from the most recent professional examination.
Pre-sleep social media duration was not independently associated with academic performance after adjustment.
Nocturnal checking was not associated with academic performance outcomes.
Results
The proportion of social media use devoted to academic work was positively associated with higher examination scores, but platform-derived purpose classification was not associated with either outcome.
Each unit increase in the proportion of social media use devoted to academic work was associated with examination scores that were 2.148 percentage points higher (B = 2.148, 95% CI 1.513 to 2.782, p < 0.001).
Classification of platform-derived purpose was not associated with either sleep quality or academic performance.
The authors noted that 'conclusions about purpose depended on how it was measured.'
The two measures of purpose (platform-derived classification vs. proportion devoted to academic work) yielded different conclusions, highlighting a methodological sensitivity.
Results
Fitting both outcomes in a single model confirmed outcome specificity of social media exposures, particularly for pre-sleep use.
The difference in standardized β for pre-sleep use between the two outcomes was -0.271 (p = 0.006), supporting outcome specificity.
The difference in standardized β for total daily use between the two outcomes was 0.235 (p = 0.023).
Outcome specificity was described as 'most clearly' demonstrated for pre-sleep use.
This dual-outcome modeling approach was used to confirm that the timing versus amount distinction was not artifactual.
Discussion
The study design and sampling method limit causal inference and generalizability beyond the participating institutions.
The authors explicitly state: 'These exploratory associations from a convenience sample cannot establish causal direction.'
Findings 'apply to the participating institutions rather than to Pakistani students generally.'
The authors call for 'replication with prospective designs and objective measures in future studies.'
Academic performance was self-reported, introducing potential for recall or social desirability bias.
What This Means
This research suggests that social media use affects sleep and academic performance through different mechanisms among medical and dental students in Pakistan. Specifically, using social media right before bed was the factor most strongly linked to worse sleep quality, while the total amount of time spent on social media each day was more strongly linked to lower exam scores. Importantly, these two outcomes were driven by different aspects of social media behavior — it wasn't simply that more social media use was bad for everything, but rather that timing mattered most for sleep and overall quantity mattered most for grades.
The study also found that how researchers measure the 'purpose' of social media use can dramatically change their conclusions. When purpose was measured by which platforms students used, it showed no association with either sleep or grades. But when purpose was measured as the share of time students spent using social media for academic work, more academic use was linked to better exam scores. This suggests that grouping social media into 'academic' or 'social' platforms may be too crude a measure, and that actual behavior on those platforms matters more.
This research suggests that interventions aimed at improving sleep in student populations might benefit from focusing specifically on reducing social media use in the period just before sleep, rather than targeting overall daily use. Conversely, efforts to support academic performance might focus on reducing total screen time or redirecting it toward educationally productive use. However, because this was a cross-sectional survey using convenience sampling at ten institutions in one Pakistani province, it cannot prove cause and effect, and the findings may not apply to all medical students in Pakistan or elsewhere. Future studies using objective measurements and following students over time are needed to confirm these patterns.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Atif K, Ahmad A, Khan A, Khan A, Ullah A, Amjad M, et al.. (2026). Differential associations of the timing and amount of social media use with sleep quality and academic performance among undergraduate medical and dental students in Pakistan: a cross-sectional study.. Frontiers in public health. https://doi.org/10.3389/fpubh.2026.1968305