Removing the Rose-tinted glasses: The effects of antidepressant drug withdrawal on mood and neurocognitive function

: (Alternative Format Thesis)

Abstract

Selective serotonin reuptake inhibitors (SSRIs) are a primary treatment for depression with the number of prescriptions rising year on year. SSRI discontinuation can present challenges for patients and clinicians alike in terms of planning the rate, timing, and duration of withdrawal, and managing the risk of side effects (like nausea and vomiting) and the risk of relapse. This thesis aimed to address and explore gaps in the antidepressant withdrawal experience by employing a mixed-methods approach, which included qualitative interviews, ecological momentary assessment (EMA) data and neurocognitive tasks. I aimed to investigate the lived experience of antidepressant withdrawal and explore its effects on mood, social interactions, and cognition. I also sought to inform theories of antidepressant action and provide insights that could inform safer withdrawal practices and enhance clinical support for individuals discontinuing antidepressants.

To address my first aim, I conducted semi-structured interviews with a community sample of participants (n = 20) who had attempted to withdraw in the past year. I found that the experience of withdrawal extends beyond physical symptoms to include profound social, emotional and cognitive consequences. Participants reported both positive and negative experiences in these domains highlighting the complexities of withdrawal beyond physical side effects or symptom re-emergence. For example, many individuals described struggling with emotional blunting, difficulties in social relationship maintenance, and cognitive suppression participants after withdrawal. Additionally, they expressed frustration over the lack of evidence based medical guidance and support for effective withdrawal, echoing concerns raised in previous research. Findings from this study underscored the importance of a more structured approach to withdrawal management and led me to investigate these effects quantitatively to assess withdrawal effects in real-time.

I then conducted a study using EMA to investigate the relationships between mood, social interactions and depressive symptoms (n = 129) which I believed would inform the development of my primary care study in later chapters. I found that while baseline depressive symptoms were negatively associated with social interaction quality there was a bi-directional relationship between reported mood and social interaction quality with positive affect predicting higher interaction quality. Interestingly, while depressive symptoms correlated with lower perceived interaction quality, there was no association with interaction quantity. Exploratory analyses showed that this relationship could be described as a time-lagged bidirectional relationship, indicating that current mood significantly influenced subsequent perceived social interaction quality ratings and vice versa. Additionally, I attempted to examine the impact of antidepressant use on these associations, though a limited number of antidepressant users in the sample restricted my ability to draw firm conclusions. Building on this study which reinforced the importance of social connectedness in mood regulation, in the second half of my thesis I investigated the effect of antidepressant withdrawal on mood (Chapter 4) and neurocognitive function (Chapter 5) in a sample of primary care patients (n= 72).

In chapter 4 I report on the longitudinal EMA study as patients attempted to withdraw (n =38) or stay on their antidepressant medication (n = 31). This was one of the first studies to track early withdrawal effects in real-time on EMA outcomes such as mood score, sleep, and withdrawal side effects. Contrary to expectations, I found little evidence that withdrawal led to significant declines in mood, increases in negative affect, or disruptions in sleep within the first four weeks of withdrawal. Furthermore, withdrawal symptoms were generally mild, and no substantial differences emerged between those who continued versus discontinued their medication. These findings contrast with retrospective reports of severe withdrawal effects and suggest that withdrawal experiences may be highly individualised, influenced by context, prior expectations, and withdrawal methods. Thus, early withdrawal may not be as disruptive as expected however longer-term effects of withdrawal remain an open question.

I further investigated the neurocognitive effects of antidepressant withdrawal in the same primary care sample. I found limited evidence that withdrawal influenced affective or social cognitive processing over time, despite using similar tasks that have previously indicated antidepressant treatment effects. However, there was some indication that withdrawal altered effort-based decision-making, with those discontinuing medication showing delayed decision times in tasks requiring cognitive effort. Importantly, neurocognitive task performance did not strongly predict changes in depressive or anxiety symptoms, suggesting that withdrawal-related cognitive shifts may not necessarily translate into clinical outcomes. These findings contribute to an emerging literature on how withdrawal may influence cognitive function, though the mechanisms remain unclear.

Taken together my findings highlight a discrepancy between the qualitative and quantitative studies of antidepressant withdrawal effects. The qualitative study revealed pronounced emotional, cognitive, and social challenges (alongside some benefits of withdrawal in these domains) while the quantitative studies indicated relatively stable mood and cognitive functioning in the early withdrawal period. This difference may be explained by differences in sampling as qualitative participants may have been more likely to participate due to having distressing withdrawal experiences while primary care participants were in remission undergoing supervised withdrawal. Important methodological considerations may also have played a role with a potential for recall biases in retrospective reports versus the advantages of real-time assessments in capturing momentary experiences which may be more nuanced.

Overall, this thesis provides novel insights into the early effects of withdrawal and highlights the complexity of experiences and the importance of acknowledging individual variability in responses and different assessment methods. These findings have important implications for clinical practice, underscoring the need for personalised withdrawal support and improved guidance for both patients and healthcare providers. Future research should continue to integrate different methodological approaches and assess effects over a longer period of time (or multiple sampling opportunities across the full withdrawal period) to better understand the full scope of antidepressant withdrawal and develop evidence-based strategies for safe discontinuation.

Date of Award	12 Nov 2025
Original language	English
Awarding Institution	University of Bath
Supervisor	Graeme Fairchild (Supervisor), Katherine Button (Supervisor), Katherine Button (Supervisor), Graeme Fairchild (Supervisor), Katherine Button (Supervisor) & Graeme Fairchild (Supervisor)