One in three students who seek help from university counselling services does not get better with the current treatment. Depression and anxiety affect roughly 12% and 11% of UK students respectively, yet the standard one-size-fits-all approach to psychological therapy fails a significant minority. Students face distinct pressures—academic stress, financial strain, and lack of social support—that may alter how they respond to treatment compared to adults. Existing studies on recovery trajectories have focused on adult or overseas populations, leaving UK students unexamined. This project will track how students’ symptoms change during therapy, grouping them by patterns of improvement or stagnation, and identify which factors—such as unemployment, suicidal thoughts, or medication use—predict who will not improve. If successful, the research could help university counselling services tailor intervention length and type to individual students, reducing the number who drop out or finish therapy without meaningful benefit. In the longer term, it may shift how mental health support is designed for young adults, moving from a uniform model to one that accounts for different recovery paths.
View original technical description
The proposed research will explore anxiety and depression profiles in students, the change of these symptoms during psychological interventions and factors that influence the changes. This analysis aims to contribute to the improvement of intervention effectiveness delivered by University counselling services (UCSs) to the student population. Depression and anxiety are the most common mental health diagnoses that students report (about 12% and 11% respectively), and the symptoms of these conditions are even more common. The student population differ from adults because of the age factor and special risks such as academic stress, lack of social support when studying away from home, and financial burden associated with tuition fees. This leads to a different response to psychological interventions. Despite university counselling services using evidence-based intervention to help students overcome mental health issues, about a third of them do not respond to the treatment. This frightening proportion calls for a more nuanced approach, which includes inter alia investigating how differently students progress during receiving interventions and which factors are associated with that progression. That can be done by comparing two models: (i) the dose-response (DR) model based on the assumption that all clients follow the same recovery trend captured by a negatively accelerating curve and (ii) the good enough level (GEL) model assuming clients tend to quit interventions when they reach good enough improvement and thus should be stratified based on the intervention duration. The latter approach got more empirical support. However, this analysis has been performed either for adults or for overseas students. To the best of my knowledge, there are no studies of UK student samples. Another way to study clients? progress is to group them based on adherence to various trajectories of changes and identify predictors of trajectory class membership. Mental health comorbidity, unemployed status, suicidality and the usage of medication received the strongest support to increase the odds of following a non-improving route for adults. At the same time, there is a lack of research done for the student population. My research aims to address these gaps by answering the key research questions: ? What is the optimal duration of psychological interventions for students with common mental health problems? ? Which depression and anxiety trajectory profiles do students adhere to while getting support from counselling services? ? Which clinical and demographic variables predict profile membership? ? How do students differ from the adult population in terms of trajectory profiles? ? How does the measure used to track the symptoms affect the results of trajectory analysis?
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know