What 602 Pakistani University Students Taught Us About Depression, Anxiety and Stress

What can 602 Pakistani university students teach us about depression, anxiety and stress? One of the clearest lessons is that student mental health cannot be reduced to the familiar assumption that medical students are always the most distressed group. In our cross-sectional study of undergraduates in Lahore, non-medical students reported higher depressive and anxiety symptoms than medical students, while the difference in stress was not statistically significant after adjustment. Just as importantly, academic discipline explained only a small part of the overall variation in mental health scores.

By Taimoor Asghar

The findings encourage a broader view of university mental health. They point toward factors such as gender, sleep and previous treatment for depression while also showing why researchers should examine the performance of the questionnaires used to measure psychological symptoms. The study therefore offers lessons not only about which students reported more symptoms, but also about how depression, anxiety and stress can be assessed and interpreted in university populations.

Why Study Depression, Anxiety and Stress in Pakistani University Students?

University life combines academic demands with major social, financial and personal transitions. Students may be adapting to a new institution, managing examinations, planning careers, living away from family or balancing education with responsibilities outside university. These pressures do not automatically produce a mental disorder, but they make student mental health an important area for research and support.

The challenge is particularly interesting when comparing academic disciplines. Medical education is frequently associated with intense workloads, competitive examinations, clinical responsibilities and exposure to illness. That can create an expectation that medical students will necessarily report the highest levels of psychological distress.

But an assumption is not the same as evidence. Students studying other subjects may face different pressures, including uncertainty about employment, academic resources, family expectations, finances or future career pathways. Comparing groups directly can therefore reveal whether the usual narrative is actually supported in a particular population.

Our study, published in BMC Psychology in August 2026, approached the question using data from 602 undergraduate students attending universities in Lahore, Pakistan. Readers who want the complete methodology and statistical results can access the published BMC Psychology study on PHQ-9 and DASS-21 symptom profiles among Pakistani university students.

Who Were the 602 Students?

The study included 602 undergraduate students, divided into two academic groups:

• 424 medical students
• 178 non-medical students

Participants completed two widely used self-report instruments: the Patient Health Questionnaire-9, commonly known as the PHQ-9, and the 21-item Depression Anxiety Stress Scales, or DASS-21.

The PHQ-9 measures depressive symptoms using nine items. It is widely used in research and healthcare settings as a depression severity measure. The DASS-21 contains three seven-item subscales intended to assess depression, anxiety and stress-related negative emotional states.

Importantly, scores on these questionnaires are not identical to a psychiatric diagnosis. Self-report instruments can identify symptom burden and patterns, but diagnosis requires appropriate clinical assessment. That distinction matters when interpreting any university mental health survey.

The Unexpected Finding: Non-Medical Students Reported More Depressive Symptoms

The comparison produced a finding that challenges a common expectation. The median PHQ-9 score was 10 among non-medical students and 9 among medical students.

The difference was not limited to an unadjusted comparison. In multivariable models reported in the study, academic discipline remained associated with PHQ-9 scores, with non-medical students showing higher depressive symptom scores. The adjusted difference was approximately 1.43 PHQ-9 points, with a reported p value of 0.004.

A similar pattern appeared on the DASS-21 depression subscale. Non-medical students had higher adjusted depression scores, with an estimated difference of approximately 1.82 points and a p value of 0.003.

This does not mean that medical students were psychologically healthy or that medical education is harmless. It means something more specific: within this sample, the data did not support the assumption that medical students had higher depressive symptom scores than their non-medical peers.

Why this distinction matters

If universities design mental health services around the idea that only traditionally high-pressure disciplines require attention, students in other faculties may be overlooked. The findings suggest that campus mental health planning should consider the wider student population rather than treating one academic discipline as a sufficient marker of vulnerability.

At the same time, the size of the difference should not be exaggerated. Statistical significance tells us that an observed association is unlikely to be explained by sampling variability alone under the assumptions of the model. It does not tell us that academic discipline is the dominant cause of students’ psychological difficulties.

In fact, the study’s regression models explained relatively little of the total variation in scores. Reported R-squared values ranged from approximately 0.029 to 0.041. In practical terms, the measured variables in these models accounted for only a small percentage of the differences between individual students.

Anxiety Showed a Similar Pattern

The comparison was not restricted to depression. After adjustment, non-medical students also reported higher DASS-21 anxiety scores than medical students. The estimated difference was approximately 2.31 points, with a reported p value of 0.001.

This reinforces the argument against assuming that psychological burden is concentrated in medical education. Different academic environments may carry different combinations of pressures, opportunities and uncertainties.

However, the study was cross-sectional. It measured participants at one period rather than following them over time. We therefore cannot conclude that being a non-medical student caused higher anxiety or depression. Academic discipline could be associated with other personal, socioeconomic, institutional or educational characteristics that were not fully captured by the analysis.

A better interpretation is that discipline was associated with symptom scores in this sample and deserves further investigation.

Stress Was Different: The Discipline Gap Was Not Statistically Significant

The stress results add an important layer of nuance. Although non-medical students showed higher adjusted depression and anxiety scores, the adjusted difference in DASS-21 stress scores was not statistically significant. The estimated difference was approximately 0.89 points, with a p value of 0.089.

This is useful because depression, anxiety and stress are related but are not necessarily interchangeable concepts. A group difference observed for one psychological domain should not automatically be assumed to exist for every other domain.

It also illustrates why research examining multiple symptom dimensions can be more informative than relying on a single headline measure. Two groups may differ in depressive symptoms while showing less evidence of separation in stress symptoms.

Gender Was More Consistently Associated With Symptoms

Academic discipline received much of the initial attention because the study directly compared medical and non-medical students. Yet one of the more consistent findings concerned gender.

Female students had higher scores across the PHQ-9 and all three DASS-21 domains in the adjusted analyses, with reported p values below 0.001.

This does not establish that gender itself biologically causes higher psychological distress. Gender-related differences in mental health can reflect a complex mixture of social expectations, exposure to stressors, opportunities, reporting patterns, structural inequalities and other factors. A cross-sectional survey cannot disentangle all of these mechanisms.

Nevertheless, the consistency of the association across depression, anxiety and stress suggests that gender should not be ignored when universities assess patterns of student distress or plan accessible mental health support.

Sleep Was Linked With Depressive Symptoms

Another practical finding concerned sleep. Each additional hour of sleep was associated with a 0.37-point lower PHQ-9 score in the adjusted model, with a reported p value of 0.010.

The result fits with the broader understanding that sleep and mental health are closely connected. However, the direction of causation cannot be determined from this study. Poor sleep could contribute to psychological symptoms, depression may disrupt sleep, or both may be influenced by other factors such as workload, physical health, lifestyle or stressful circumstances.

That distinction is especially important in cross-sectional research. An association between two variables should not be converted into a treatment claim.

Still, sleep is relevant to campus health because university schedules and student routines can influence opportunities for sufficient rest. Universities considering mental health promotion may therefore benefit from looking beyond counselling services alone and examining the broader conditions that shape student well-being.

Previous Depression Treatment Was a Strong Predictor

Among the variables examined, prior treatment for depression showed a particularly strong association with current depressive symptom scores. The study reported an adjusted coefficient of 4.15 on the PHQ-9, with a p value below 0.001.

This is clinically understandable without implying causality. A history of treatment can identify students who have previously experienced significant depressive symptoms or other mental health concerns. Some may continue to experience symptoms, while others may experience recurrence.

The finding highlights why mental health history can be more informative than broad labels such as faculty or degree programme. Two students studying the same subject may have very different mental health trajectories.

It also supports a broader principle: effective student support systems need to be capable of responding to individual needs rather than relying solely on demographic or academic categories to determine who deserves attention.

The Study Also Asked Whether Our Measurement Tools Were Working Well

A distinctive feature of this research was that it did more than compare average scores. We also examined the psychometric performance of the PHQ-9 and DASS-21.

This matters because conclusions about mental health are only as trustworthy as the measurement process supporting them. A questionnaire developed or validated in one population should not automatically be assumed to behave identically in every cultural, linguistic or educational context.

Reliability was acceptable

The reported Cronbach’s alpha coefficients across the scales ranged from 0.82 to 0.88. These values indicated good internal consistency in this sample.

Internal consistency asks whether items intended to measure a related construct tend to produce coherent responses. It does not prove that a questionnaire diagnoses depression or anxiety correctly, but it is an important component of psychometric evaluation.

The DASS-21 factor structure showed adequate fit

Confirmatory factor analysis of the DASS-21 produced an adequate model fit. The reported comparative fit index was 0.954, the Tucker-Lewis index was 0.948, the root mean square error of approximation was 0.052, and the standardized root mean square residual was 0.031.

These statistics support the overall measurement structure, but another result deserves attention: correlations between some latent DASS-21 factors were extremely high. Depression and stress correlated at 0.939, while anxiety and stress correlated at 0.949.

Such high correlations suggest substantial overlap between the constructs as measured in this population. The authors therefore raised a reasonable psychometric question about how distinctly the three DASS-21 subscales operate among these students.

This does not make the DASS-21 useless. Rather, it demonstrates why researchers should inspect how a scale functions instead of assuming that three named subscales automatically represent three sharply separated psychological dimensions.

Some Symptoms Carried More Information Than Others

The study also used item response theory, a statistical approach that examines how individual questionnaire items perform across different levels of an underlying trait.

Several symptoms were particularly informative. Items involving self-worth, concentration, feeling down and appetite showed comparatively high discrimination. In contrast, the PHQ-9 item relating to diminished interest or pleasure showed the lowest discrimination parameter in the reported analysis, with an estimated value of 0.468.

These findings should not be interpreted as instructions to ignore any particular symptom. Every PHQ-9 item contributes to the established questionnaire, and clinical assessment should never be reduced to whichever items happen to have the strongest statistical properties in one sample.

For researchers, however, item-level findings can help illuminate which questions best distinguish between students at different levels of depressive symptom burden in a particular population.

Network Analysis Highlighted Self-Worth and Cognitive Symptoms

Another analysis treated individual symptoms as an interconnected network rather than looking only at total questionnaire scores.

Self-worth, concentration and downheartedness emerged among the most central symptoms in the estimated network. Conceptually, network analysis asks how symptoms relate to one another and which symptoms occupy relatively connected positions within the observed statistical structure.

That can be useful for hypothesis generation. For example, a central symptom might deserve further study as a potential marker of wider psychological difficulty.

But centrality in a cross-sectional symptom network does not demonstrate that changing one symptom will necessarily cause improvement in others. The study’s network stability coefficients ranged from 0.31 to 0.44, reinforcing the need for appropriate caution when interpreting rankings of central symptoms.

Longitudinal studies, repeated measurements and intervention research would be needed before making stronger claims about symptom pathways or treatment targets.

Medical and Non-Medical Students Appeared to Be Measured Comparably

The study also assessed measurement invariance across academic discipline. Support for invariance is important when comparing groups because researchers need reasonable confidence that a questionnaire is measuring the underlying construct in a comparable manner in both groups.

If the same questionnaire functioned fundamentally differently for medical and non-medical students, differences in total scores could partly reflect measurement artefacts instead of true differences in reported symptoms.

The reported support for measurement invariance therefore strengthens the interpretability of the discipline comparisons, while still leaving all the usual limitations of observational research in place.

What the Study Does Not Tell Us

The findings are useful, but their boundaries are equally important.

It cannot prove causation

Because this was a cross-sectional study, exposure and outcome variables were measured within the same general study period. The data cannot establish whether academic discipline, sleep or another associated factor caused changes in mental health.

It should not be generalized to every Pakistani student

The participants were undergraduates recruited from universities in Lahore. Pakistan contains substantial diversity across provinces, cities, rural areas, educational systems, socioeconomic circumstances and types of institution. A sample of 602 students provides valuable evidence, but it is not a census of Pakistani university students.

Questionnaires are not psychiatric diagnoses

The PHQ-9 and DASS-21 measure self-reported symptoms. High scores can indicate meaningful psychological distress and may justify further assessment, but they should not be treated as substitutes for a diagnostic interview conducted by an appropriately qualified professional.

The statistical models explained only a small proportion of individual differences

The low R-squared values are a reminder that mental health is multifactorial. Academic discipline, gender, sleep and recorded clinical history do not capture everything influencing an individual’s emotional state.

What Universities Can Learn From These 602 Students

The study does not prescribe a single intervention, but several practical lessons emerge from the evidence.

• Mental health programmes should be university-wide. Non-medical students should not be assumed to face little psychological burden simply because medical education has a reputation for being stressful.
• Support should be based on need rather than stereotypes. Academic discipline accounted for only a small part of the variation between students.
• Gender differences deserve further investigation. Female students reported higher symptom scores across multiple domains in the adjusted analyses.
• Sleep belongs in conversations about student well-being. The observed association with depressive symptoms does not establish causation, but it supports further attention to sleep health in university populations.
• Students with previous mental health difficulties may need continuity of support. Prior depression treatment was strongly associated with current PHQ-9 scores in this study.
• Measurement quality matters. Universities and researchers using screening tools should understand what those tools measure, their limitations and whether they perform appropriately in the population being assessed.

The Bigger Lesson: Look Beyond the Medical vs Non-Medical Debate

Perhaps the most important lesson from these 602 students is that the question should not end with, “Which faculty is more depressed?”

Yes, the comparison produced a noteworthy result: non-medical students reported higher depression and anxiety scores than medical students in this Lahore sample. But the study simultaneously showed why that result should not become a new stereotype replacing the old one.

Discipline explained relatively little of the variation in psychological symptoms. Gender, sleep and previous depression treatment were also associated with outcomes, and many other influences were inevitably outside the scope of the models.

Student mental health is therefore better understood as a multidimensional issue than as a competition between academic programmes.

The psychometric findings add another layer to that lesson. Depression, anxiety and stress are not simply numbers waiting to be collected. How we measure them matters. The substantial correlations among DASS-21 dimensions, differences in item discrimination and symptom-network findings demonstrate that mental health questionnaires contain structure that can be explored rather than treated as unquestionable black boxes.

Where Research Should Go Next

Future research could build on these findings in several ways. Studies involving universities across different Pakistani regions would help determine whether similar discipline differences appear elsewhere. Longitudinal studies could follow students through academic years to explore whether changes in workload, sleep, examinations or life circumstances precede changes in symptoms.

Researchers could also examine socioeconomic circumstances, financial pressure, academic performance, social support, living arrangements, physical health, substance use and other potentially relevant factors. Qualitative interviews could complement questionnaire scores by showing how students themselves describe the sources and meanings of their distress.

Repeated psychometric evaluation would also be valuable. Measurement tools should continue to be tested across languages, disciplines and demographic groups rather than assuming that a scale performs identically in every setting.

Most importantly, future studies should resist oversimplification. A statistically significant difference between two groups may be scientifically interesting without defining every individual within those groups.

Final Takeaway

The study of 602 Pakistani university students challenges a familiar assumption. In this sample, non-medical students reported higher depressive and anxiety symptoms than medical students, while adjusted stress scores did not differ significantly by discipline. Female students reported higher scores across the measured domains, more sleep was associated with lower PHQ-9 scores, and previous depression treatment showed a strong relationship with current depressive symptoms.

At the same time, academic discipline explained only a small proportion of overall variation. That may be the most useful message for universities, researchers and students: mental health cannot be understood from someone’s degree programme alone.

Good student mental health research requires both broad support for people experiencing distress and careful measurement of the symptoms being studied. Rather than asking which group deserves attention, universities should work toward systems capable of recognizing and responding to mental health needs wherever they occur.

Medical disclaimer: This article is for educational and research communication purposes only. The PHQ-9 and DASS-21 are symptom-assessment instruments and information discussed here should not be used to diagnose yourself or another person. Anyone experiencing persistent depression, severe anxiety, significant impairment, thoughts of self-harm or other concerning mental health symptoms should seek assessment from an appropriately qualified healthcare or mental health professional. Urgent or emergency symptoms require prompt local professional assistance.