Academic Major Explained Only a Fraction of Depression Differences—So What Matters More?

Academic major may be associated with depressive symptoms, but it is unlikely to explain most of the mental-health differences between university students. In our 2026 study of 602 undergraduates in Lahore, non-medical students reported somewhat higher depression scores than medical students, yet the statistical models explained only a small fraction of the overall variation. The more useful conclusion is therefore not that one academic discipline is mentally healthier than another. It is that student depression is shaped by a broader combination of personal history, sleep, gender-related or social influences, and factors that were not captured by academic major alone.

By Taimoor Asghar

Academic major and depression: what did the study actually find?

The study compared 424 medical students with 178 non-medical students from universities in Lahore, Pakistan. Participants completed the Patient Health Questionnaire-9 (PHQ-9) and the Depression Anxiety Stress Scales-21 (DASS-21), and the analysis examined differences between academic groups alongside several other variables.

The central finding was somewhat counterintuitive. Non-medical students had a median PHQ-9 score of 10 compared with 9 among medical students. After adjustment for the other variables included in the regression model, academic discipline remained associated with PHQ-9 scores, with non-medical students reporting higher depressive symptoms. Similar adjusted differences appeared for DASS-21 depression and anxiety, whereas the difference in DASS-21 stress was not statistically significant.

Those results matter because medical students are frequently discussed as a uniquely vulnerable university population. Medical education undoubtedly contains substantial academic and professional pressures, but the findings show why that observation should not become an assumption that medical students must always have worse mental health than students in other disciplines.

The full study, Psychometric properties and symptom profiles of the PHQ-9 and DASS-21 among medical and non-medical university students, was published in BMC Psychology in August 2026.

The most revealing result was the small amount of variance explained

A statistically significant difference between two groups does not necessarily mean that group membership explains much of what is happening at the individual level. This distinction is especially important when interpreting mental-health research.

In the study, the regression models had R-squared values of approximately 0.029 to 0.041. In practical terms, the variables entered into those models collectively accounted for only about 3% to 4% of the variation in symptom scores, depending on the outcome.

That does not make the observed associations meaningless. It changes the question we should ask.

Instead of asking, “Which academic group is more depressed?” a better question is, “Why do students within the same academic group differ so much from one another?”

Two students can attend the same type of university program while having very different sleep patterns, financial pressures, family circumstances, social support, prior mental-health histories, academic expectations, relationships, living conditions and coping resources. A label such as “medical student” or “non-medical student” cannot capture all of that variation.

What appeared to matter more than academic major?

The study cannot establish a definitive hierarchy of causes because it was cross-sectional and observational. It can, however, show which measured characteristics were associated with symptom scores after adjustment and which findings deserve further investigation.

1. Previous treatment for depression was a much stronger marker

Among the variables examined in the PHQ-9 model, prior treatment for depression showed the largest reported coefficient. Students with a history of depression treatment scored about 4.15 points higher on the PHQ-9 after adjustment for the variables included in the model.

This finding should be interpreted carefully. It does not mean that treatment causes higher depression scores. A far more plausible interpretation is that a history of treatment identifies students who have previously experienced clinically important depressive symptoms or related difficulties. Some may continue to experience symptoms, while others may have recurrent episodes or residual symptoms.

The distinction illustrates a broader principle in observational research: a predictor can be highly informative without being causal.

For universities, mental-health history may therefore be more relevant to understanding individual vulnerability than academic discipline alone. At the same time, universities should avoid treating students with previous mental-health care as a homogeneous high-risk category. A treatment history tells us something about the past; it does not determine an individual’s present condition or future course.

2. Sleep duration showed an independent association

Sleep was another notable finding. Each additional reported hour of sleep was associated with a 0.37-point lower PHQ-9 score in the adjusted analysis.

This association is clinically and practically interesting because sleep and mood are closely intertwined. However, the cross-sectional design prevents us from determining direction. Shorter sleep could contribute to worse mood, depressive symptoms could disrupt or shorten sleep, or both could reflect another underlying stressor.

There is also a difference between sleep duration and sleep quality. Simply spending more time in bed does not guarantee restorative sleep, and the study did not establish that increasing sleep by a particular amount would produce a corresponding reduction in depression scores.

Still, sleep deserves more attention in university mental-health research because it cuts across academic disciplines. Early classes, late-night studying, shift work, commuting, digital media use, examination periods, irregular schedules and social pressures can all alter students’ sleep routines.

This makes sleep a potentially more useful campus-wide target for further research and health promotion than assuming psychological distress belongs mainly to one faculty.

3. Female students reported higher symptoms across the measured domains

Female gender was associated with higher scores across the PHQ-9 and all three DASS-21 domains in the adjusted analyses reported in the study.

Again, an association does not tell us why the difference exists. Gender-related differences in reported psychological symptoms may reflect multiple interacting influences, including social circumstances, exposure to stressors, cultural expectations, differences in symptom recognition or disclosure, biological factors, safety concerns and access to support.

It would therefore be inappropriate to reduce this result to the idea that gender itself directly produces depression. Instead, the finding suggests that researchers examining university mental health in Pakistan should investigate the mechanisms underlying these differences rather than treating gender merely as a demographic adjustment variable.

Universities considering mental-health services should also recognize that population-wide provision and attention to potentially vulnerable groups are not mutually exclusive. A service can be available to all students while still examining whether particular groups face distinct barriers or stressors.

Most of the difference remained unexplained

Perhaps the most consequential implication of the low R-squared values is what the study did not explain.

If the models accounted for only a small percentage of variation in symptom scores, then the overwhelming majority of differences between students were associated with factors outside the measured model, random variation, measurement limitations or combinations of influences that were not captured adequately.

Potential areas for future investigation include financial strain, family relationships, loneliness, social support, academic performance, examination pressure, commuting burden, housing conditions, chronic illness, substance use, bullying or harassment, relationship difficulties, employment alongside study, perceived career uncertainty, physical activity, sleep quality and stressful life events.

These are research priorities rather than conclusions from the present dataset. The study did not demonstrate that any particular unmeasured factor explains the remaining variance.

That distinction matters. When a statistical model leaves most variation unexplained, the scientifically responsible response is not to fill the gap with plausible-sounding assumptions. It is to design better studies.

Why a significant academic-major effect can still be small

Readers sometimes interpret a p-value below 0.05 as evidence that a variable is a major determinant of an outcome. Statistical significance does not tell us that.

A group difference can be statistically detectable while explaining little of the variation between individuals. Imagine two university faculties whose average depression scores differ modestly. The distributions can still overlap substantially: many students in the group with the lower average score may have more symptoms than many students in the group with the higher average.

This is one reason individual screening or clinical assessment should never be based on a person’s academic major.

The finding also cautions against headlines suggesting that one type of student is “more depressed” as though faculty membership were a sufficient explanation. The observed difference is a population-level comparison within a specific sample, not a diagnostic rule.

Depression scores are not the same as depression diagnoses

The PHQ-9 is a nine-item questionnaire assessing the frequency of depressive symptoms. It is widely used as a screening and symptom-severity instrument, but a questionnaire score by itself is not equivalent to a comprehensive clinical diagnosis.

This distinction is particularly important when discussing university surveys, where researchers often collect self-reported measures from hundreds of participants without conducting individual diagnostic interviews.

A student may report a high symptom score for several reasons, and interpreting that result properly may require information about duration, impairment, medical conditions, medications, substance use, bereavement, other psychiatric symptoms and personal circumstances.

The study therefore provides evidence about reported depressive symptoms, not proof that a particular percentage of either academic group had major depressive disorder.

The symptom pattern may be as informative as the total score

The investigation went beyond comparing average or median scores. Item response theory and network analysis were used to examine how individual symptoms behaved within the PHQ-9 and DASS-21.

Several PHQ-9 symptoms, particularly self-worth, concentration, depressed mood and appetite-related symptoms, showed relatively strong discrimination in the item-response analysis. The network analysis also identified self-worth and concentration among the more central symptoms, although the reported network stability coefficients were moderate and the findings should not be interpreted as establishing causal symptom pathways.

This reinforces another reason academic major cannot tell the whole story. Two students with similar PHQ-9 totals can arrive at those scores through different symptom combinations. One may be dominated by concentration difficulties and feelings of worthlessness, while another may report sleep disruption, fatigue and loss of interest.

For research, examining symptoms rather than totals alone may reveal patterns hidden by a single composite score. In clinical practice, however, network centrality or item discrimination should not replace an individualized assessment.

What should universities take from these findings?

The most defensible implication is that campus mental-health strategies should not be designed around the assumption that distress belongs primarily to one prestigious, difficult or traditionally high-pressure academic program.

A broader approach could include:

• making mental-health information and support visible across faculties rather than concentrating exclusively on selected disciplines;
• paying attention to sleep and scheduling as components of student well-being;
• ensuring students with previous mental-health difficulties know how to access appropriate continuing support;
• investigating barriers experienced by female students and other potentially vulnerable groups;
• training staff to recognize meaningful changes in functioning rather than relying on stereotypes about which students are expected to struggle;
• collecting better longitudinal data on financial, social, academic and personal stressors.

These measures should complement, not replace, appropriate professional mental-health care.

What should researchers study next?

The next step is not simply to repeat medical-versus-non-medical comparisons in new samples. Future work can improve our understanding by measuring the mechanisms that may sit beneath those categories.

Longitudinal studies would be especially valuable. Following students over semesters could help determine whether changes in sleep, financial stress, examinations, social support or academic workload precede changes in depressive symptoms. Repeated measurement would also help distinguish persistent vulnerability from temporary fluctuations around examinations or other stressful periods.

Researchers could additionally examine whether associations differ by academic year, socioeconomic position, living arrangement, employment status, relationship status or specific degree program. “Non-medical” itself contains enormous diversity: engineering, business, humanities, social sciences and other disciplines may involve very different environments.

Qualitative research could add another layer by asking students what they themselves believe drives periods of psychological distress. Statistical models identify patterns, but interviews and focus groups can illuminate experiences that a fixed questionnaire may overlook.

Avoid replacing one stereotype with another

Finding higher symptom scores among non-medical students in this sample should not lead to a new stereotype that non-medical students are generally at greater risk than medical students.

The study was cross-sectional, involved students from Lahore, used self-report measures and included unequal numbers of medical and non-medical participants. Its results should be interpreted within that context rather than generalized automatically to every university in Pakistan or internationally.

The stronger message is that academic discipline is an incomplete proxy for the conditions shaping student mental health.

Medical students can experience serious depression. Non-medical students can experience serious depression. Students in either group can also have few or no depressive symptoms. The considerable variation within academic groups is precisely why population averages should not become assumptions about individuals.

The question should shift from “Which major?” to “Which circumstances?”

Academic-major comparisons are attractive because they are easy to understand. They divide a complicated population into recognizable groups and produce a straightforward difference. Mental health is rarely that simple.

Our results detected a difference by academic discipline, but the low explanatory power of the models shows how much information is lost when we focus on that comparison alone. Previous depression treatment, gender and sleep were meaningful correlates within the measured model, while most individual variation remained unexplained.

For student mental-health research, the more productive direction is therefore to investigate the circumstances surrounding students: their histories, daily routines, resources, pressures, relationships and changing environments.

Academic major may be one piece of the picture. It should not be mistaken for the picture itself.

Frequently asked questions

Did non-medical students have higher depression scores than medical students?

Yes. In this sample of 602 university students in Lahore, non-medical students had a higher median PHQ-9 score, and an adjusted difference remained in the regression analysis. The finding applies to this study population and should not be interpreted as proof that non-medical students everywhere have worse mental health.

Does academic major cause depression?

No causal conclusion can be drawn from this study. It was cross-sectional, meaning exposures and symptoms were assessed at approximately the same point in time. Academic discipline was associated with some symptom outcomes, but the design cannot establish that studying a particular subject caused those symptoms.

What factor had the strongest association with PHQ-9 scores?

Among the measured variables reported in the model, previous treatment for depression had the largest coefficient, with an adjusted difference of approximately 4.15 PHQ-9 points. This should be interpreted as a marker of mental-health history rather than evidence that treatment increases depression.

Was sleep related to depressive symptoms?

Yes. Each additional reported hour of sleep was associated with a 0.37-point lower PHQ-9 score after adjustment. Because the study was cross-sectional, it cannot determine whether shorter sleep contributed to symptoms, symptoms affected sleep, or other factors influenced both.

Can a PHQ-9 score diagnose depression?

The PHQ-9 is a validated screening and symptom-severity questionnaire, but a questionnaire result is not a substitute for a full professional assessment when diagnosis or treatment decisions are required.

Medical disclaimer: This article is for educational and research communication purposes only. It does not provide an individual diagnosis or treatment recommendation. Anyone experiencing persistent depressive symptoms, significant impairment, thoughts of self-harm or other serious mental-health concerns should seek assessment from an appropriately qualified healthcare professional or urgent local assistance when necessary.