Depression in Medical Students: Weaknesses of a Longitudinal Study

Depression in Medical Students: Weaknesses of a Longitudinal Study

  • Institutional Support

The institutions supporting this research include the University of Minho School of Medicine and PT Government Associate Laboratory, both located in Braga, Portugal. The authors stated that they had no sources of funding for this research (p. 8).

The explicit purpose of this research was “to contribute to the comprehension of depression in medical students, with the certainty that this knowledge will be useful to aid in the design of preventive strategies and more effective interventions to improve the quality of medical education” (p. 2). Specifically, the study aimed to determine the prevalence of depression, how that prevalence changed over the course of a medical degree program, whether the depression is persistent, what personal and academic factors might be associated with persistence or recovery, and whether those factors change over time.

Considering that the research was supported primarily by a medical school, it is likely that it had the implicit purpose of discovering interventions that could help prevent medical students from burning out and dropping out of medical school. This is supported by the lengthy discussion of the well-being of medical students and the consequences of medical student depression in the paper’s introduction. Notably, the introduction focuses on factors that are dependent on the medical student, and does not discuss any factors specific to medical schools that might cause depression and burnout in medical students. This could be an example of how the institutional support influenced the research: the medical perspective is about finding a way to treat the problem, rather than taking responsibility for, and addressing the cause.

  • Research Design

This study uses the research design of “an observational, longitudinal, and prospective study” (p. 2). The study is observational in the sense that it does not have a control group that will be examined under normal conditions and an experimental group that will be examined under altered conditions. It is longitudinal because it follows the same set of subjects over a long period of time, as opposed to a cross-sectional study which would look at a set of subjects at only one point in time. Lastly, it is a prospective study because it examines the subjects before they are known to have developed the condition of interest, depression.

This design has several major benefits. Because subjects are tracked over time starting when they are relatively healthy, the researchers may be able to use their findings to isolate risk factors for depression, and they do not need to rely on subjects to remember what happened to them years ago. Because longitudinal studies function as a series of cross-sectional studies, they share the strengths of cross-sectional studies in measuring prevalence, one of the explicit purposes of this research.

At the same time, because it is harder to get participants to participate for the entire study, the resulting data was not complete. A longitudinal study’s findings are weaker if some of the cross-sections have missing data. This research also lacked a sample of non-medical students, which would have been a useful comparison and control group.

  • Selection Strategy

The researchers selected “all, including foreign, students in the 2009–2010, 2010–2011, 2011–2012, and 2012–2013 cohorts of medical students at Medical School – University of Minho… without incentives and there were no exclusion criteria” (pp. 2-3). This selection strategy was non-random. Its limitations depend on the perspective taken on what the sample represents. If one views the sample as the entire population of medical students at the University of Minho, problems with sampling become irrelevant, because the study is not looking at just a sample of the population. On the other hand, if the population of medical students at the University of Minho is a sample of medical students worldwide, then this selection strategy would introduce selection bias because it only uses subjects from one location. It could even be viewed as a convenience sample, if not for its size (2234 participants, 238 that participated every year). From this perspective, the research has limited usefulness when applied to other contexts, unless the researchers could find a way to control for the factors that are specific to the University of Minho.

  • Key Terms

The key terms used in this research were depression, anxiety, and burnout. Depression was not conceptualized directly, and was instead operationalized by “using the BDI [Beck Depression Inventory]… one of the most used psychometric scales for assessing the severity of depression” (p. 3). Similarly, anxiety was not conceptualized, only operationalized using the “subscale Trait Anxiety Scale (T-Anxiety) of the STAI [State Trait Anxiety Inventory]” (p. 3). Burnout was briefly conceptualized as having “three dimensions: emotional exhaustion, cynicism and academic inefficiency” (p. 3), and was then operationalized using the “Maslach Burnout Inventory – student version” (p. 3).

The lack of conceptualization is not a major issue, since depression, anxiety, and burnout are very common concepts in psychology and medical research. However, the researchers do not explain why they chose these specific instruments to operationalize these key terms. In the paper’s Methods section, the validity and reliability of the depression and anxiety instruments are not mentioned, let alone discussed. The researchers explain the scoring systems and comment that the instruments have been “adapted to the Portuguese population” (p. 3), but only comment on the burnout instrument: “the version used for this study was translated into Portuguese by Faria et al. with high reliability” (p. 3). As such, without examining the researchers’ sources for these instruments, it is not clear that their chosen methods are valid and reliable.

  • Research Implementation

The research was implemented using “printed copies of the questionnaires” (p. 3), or in other words, paper surveys. Surveys were given to all participating students by the senior year students, along with “confidentiality disclaimers” (p. 3). This was repeated each October for a total of 4 years. The surveys were “completed anonymously and returned in ballots” (p. 3). The students were also given an individual identification code to link their later responses together, enabling them to perform a longitudinal analysis. The authors state that this time of the academic year was chosen because it did not have “stressful study periods or examinations” (p. 3).

The main limitation of this implementation was that the research relies on self-reported instruments. The study did not have more objective sources of data, such as clinical observation or structured interviews. Self-reported data is less reliable because of the potential for reports that are based on incorrect memories or exaggerations. While the longitudinal design should reduce the effect of flawed memory on the student responses, because the surveys were only given once per year, students still had to try to remember things that might have happened to them up to one year ago. This is a long time, especially for medical students, who will spend most of that time studying intensely.

The researchers also found that “some datasets could not be matched as they were identified by incorrect individual identification codes” (p. 8). In other words, some of the cross-sections that should have been linked together became scattered. A more reliable method of linking responses could have reduced the impact of this problem and improved the results of the research. One option that is probably effective would be to implement a double-blind system, where one group of researchers knows which students completed which responses and can prevent students from losing their past responses, while the other group of researchers can access the content of those responses but not the identity of the student who provided the response. Since the first group would not have access to the content of the responses, this method should be able to maintain anonymity in the survey responses, not just confidentiality.

In this study, “the need for informed consent was waived as part of this research” (p. 8). This makes sense in the context of the study. While survey questions can sometimes remind the research participants of past traumas and cause some harm, medical students are expected to be relatively well-informed on these subjects and may even have more direct experience through their classes and case studies. The study did not change any conditions for the participants, leaving only the small potential for harm from the surveys. As such, this research did not have any significant ethical issues.


  • Conclusion

This research has several important strengths, such as its prospective longitudinal design and anonymous, low-stress implementation. However, it also has numerous weaknesses. The sample was selected from a single school, the data was collected from self-report surveys, there was no control group of non-medical students, some longitudinal data was lost, and there was no significant discussion of factors specific to medical schools that might contribute to depression, anxiety, and burnout. Overall, I do not think it is a reliable source of knowledge.

Our psychology writing team offer a wide range of services which are included but not limited to:
– Psychology Coursework Writing Help
– Psychology Dissertation Writing Service
– Psychology Essay Writing Service
– Psychology Dissertation Help
– Psychology Assignment Writing Service
– Psychology Essay Writers