The forced swim test: Giving up on behavioral despair (Commentary on Molendijk & de Kloet, 2021).

Abstract

Depression is a prevalent, debilitating mental illness which affects over 300 million people worldwide and is often comorbid with chronic illnesses or anxiety disorders (Ménard et al., 2016; World Health Organization (WHO), 2017). Characterized by a wide range of symptoms involving disruption of mood and cognition, depression is a leading cause of disability and contributor to economic burden (Ménard et al., 2016; World Health Organization (WHO), 2017). Even as incidence of depression steadily escalates (Akil et al., 2018), research efforts to understand its increasingly complex pathology and produce more effective treatments are stymied. The symptoms of depression are heterogeneous and do not point to a single etiological origin (Nestler & Hyman, 2010). While animal models of depression and their accompanying behavioral validations are fundamental to depression research, they can also be a source of frustration. The Forced Swim Test (FST, also known as the Porsolt swim test) in particular has proven controversial. Developed by Roger Porsolt in the 1970s as a rapid behavioral screen for antidepressant compounds (Porsolt et al., 1977, 1978), the FST is now considered by some as “the gold standard animal test for depression” (Unal & Canbeyli, 2019). The classic FST is a 2-day test where a rodent is placed in an inescapable container of water and must swim for 15 min on the first day and 5 or 6 min on the second “test” day twenty-four hours later (Porsolt et al., 1977, 1978; Yankelevitch-Yahav et al., 2015; Figure 1). During the test session, animals quickly begin to float, exhibiting immobility behavior that has been interpreted as behavioral despair, learned helplessness, passive coping, psychomotor retardation, anxiety, and even autism. But what is the FST really measuring? Does immobility in this test really correspond to behavioral despair, serving as a reliable measure of depressive-like behavior? These are the questions that Molendijk and de Kloet thoroughly explore in their latest review in the Special Issue of European Journal of Neuroscience. Having followed the FST usage and interpretations for at least the past 6 years, Molendijk and de Kloet are key contributors in promoting discussions surrounding the interpretation of the FST. The current review emphasizes their previous work and continues to track FST interpretations in the field. Additionally, they further expand on usage of the FST and potential mechanisms behind its observed behaviors. Overall, the article provides unique and detailed perspectives using meta-analyses of recent FST literature. The authors begin by providing an update on the trends in the usage and interpretation of the FST since their previous assessments in 2015 (Molendijk & de Kloet, 2015) and 2019 (Molendijk & de Kloet, 2019). Molendijk and de Kloet compiled and analyzed datasets of recent publications between 2018 and 2020 which use the FST, describing updated statistical assessments of trends in usage and interpretations over time, across countries, in different journals, and between different behavioral outcomes. Here, they found that the use of the FST in publications is no longer increasing, indicating a shift in trend for the first time since the FST's inception. Pooling from a random selection of articles, the authors report a decline in interpreting FST immobility as depressive-like, particularly from European and North American regions. Scoring immobility as coping or “other” exhibited a slight increase, with a similar relationship to geographical region (Molendijk & de Kloet, 2019). Molendijk and de Kloet's analysis takes the pulse of the global scientific community's usage of the FST, allowing us to make predictions about the test's future in depression research. Molendijk and de Kloet then assembled a second dataset of FST literature from five behaviorally focused journals with large amounts of relevant publications to determine whether studies measuring different behaviors observe overlap in their outcomes. Frequently, individual studies will use a series of different behavioral tests in combination as to provide a broader picture of an animal's or group's behavioral phenotype (Molendijk & Kloet, 2021; Nestler & Hyman, 2010). Thus, this review analyzed overlap between use of the Tail Suspension Test, the Sucrose Preference Test, and anxiety-like measures (Elevated Plus Maze, Light-Dark Box, Open Field Test), in conjunction with the FST. These data provide fascinating insight into individual variability in behavior within groups. That is, an anhedonic phenotype in the Sucrose Preference Test will not always positively correlate with passive coping in the FST or anxiety-like behaviors in the Elevated Plus Maze, and vice versa. While it is convenient to assume that, for example, a chronically stressed animal will present with depressive- and anxiety-like behaviors across each test, this is certainly not the outcome in many cases. Importantly, this analysis can also inform our understanding of different interpretations of immobility. For instance, the lack of correlation between anxiety-like behaviors in several different tests and FST immobility suggests that the FST is not measuring anxious modalities. Another original aspect of the current review lies in the section where the authors gathered opinions from the editors of previously mentioned journals which frequently publish FST-related articles. Here, they asked for a point of view from each editor on submissions which describe FST immobility as depressive-like behavior. While one journal provided no opinion, most editors tend to give their trusted reviewers the larger say in whether submissions are reaching appropriate conclusions. However, most of the editors also provided opinions reflecting the idea that the FST and other behavioral outcomes should be interpreted with caution and not be relied on too greatly. Editors-in-Chief Dr. Heuser and Dr. Dantzer of Psychoneuroendocrinology specifically emphasize that authors should examine behavioral guidelines put forth by the National Institute of Mental Health (NIMH). Further on in the review, Molendijk and de Kloet address these guidelines and discuss the NIMH Research Domain Criteria (RDoC). These criteria call researchers to consider multiple levels of disease complexity (such as genetics, behavior, or circuitry) when studying aspects of psychiatric disorders, with an overarching goal of discovering mechanisms which may translate to human conditions. Importantly, we should recognize that we are individually studying aspects of a more dimensional problem, rather than “models” which tend to reduce multifaceted psychiatric disorders to a single behavioral or molecular difference. As higher governing institutions and many well-respected journals are moving towards these ideas, reviewers and authors might also become more deliberate in their experimental analyses and interpretations. With these considerations in mind, we can continue our research with the knowledge that we are contributing to a greater mission of integrating relevant information in the psychiatric field. The authors finally review the literature, providing a thorough background of interpretations of the FST over time, including its limited use as a measure for validation of antidepressant compounds. They discuss and contrast the role of the HPA-axis, including both glucocorticoid and mineralocorticoid receptors (GRs and MRs), in stress coping behaviors. Finally, they elegantly highlight and synthesize recent literature which suggests circuitry involved in coping strategy, insinuating that interactions between these “bottom-up endocrine” and “top-down circuitry” processes are involved in immobility behavioral output (i.e., coping strategy), linking these interactions to possible genetic phenotypic differences in different strains. Overall, these findings provide strong evidence for alternative explanations underlying animal performance in the FST. While controversies involving the FST are prevalent both within science and in the public, contemplating both value and ethicality, it is important to consider the following questions: Are current behavioral paradigms providing us with the validity we actually need? Is the FST being used in the correct manner to provide the most valuable information? Furthermore, are we considering these same issues for other behavioral paradigms? Are over-reliant interpretations of these behaviors restricting progress in the field of mental health? Indications in this review and elsewhere suggest that the FST is no longer considered the miracle test to predictively validate antidepressant effects, and rightly so. Perhaps, it is worthwhile to consider other approaches to evaluate animal phenotypes and validate antidepressant treatments in ways that more accurately represent the heterogeneity of clinical depression. More importantly, it is crucial for researchers to provide arguments of validity in publications, describing both strengths and weaknesses of their models so that they can be properly evaluated. In a larger sense, this review reminds us of the imperative to continue such evaluation—that we must critically assess how we are interpreting complex animal behaviors to ensure we are not overly interpreting behaviors or potentially missing out on alternative and fruitful directions. This work was supported by an NIH grant P20GM103641 as well as intramural funding from the University of South Carolina. The authors have no conflicts to declare. The peer review history for this article is available at https://publons.com/publon/10.1111/ejn.15270.