Experiencing stressful life events can increase the risk of depression. This causal relationship is the foundation of the chronic unpredictable stress model, one of the most popular animal models for studying depression. This model aims to make rats depressed by exposing them to different stressors every day for weeks or months. Some common stressors in the model include food and water deprivation, restricted movement, social isolation, or crowding. These stressors can be combined in different ways for varied periods of time and can add up to be severe from an animal welfare point of view.

One focus of the model is anhedonia, which is a state in which one cannot feel pleasure from things that would normally bring enjoyment. This is one of the main symptoms of depression. Since rats cannot
communicate how much they enjoy something like humans can, researchers have developed tests to infer the rats’ capacity for pleasure by observing their behaviour. The most popular test is the sucrose preference test, which measures how much a sugary solution is preferred over water. Here, it is assumed that the rats will naturally prefer the sweet solution because sweet food is gratifying to eat. However, this
is an indirect way of measuring pleasure in rats. An alternative test involves implanting an electrode in the rat’s brain that delivers an electrical stimulation to parts involved in pleasure. This is called intracranial self-stimulation (ICSS) because rats are taught how to deliver this stimulation themselves. The goal of the test is to determine the level of stimulation at which the rat starts to experience pleasureand act to get more of it (self-stimulate). Since the brain is directly stimulated, this test is considered a more direct approach to measuring pleasure in rats.

If a rat becomes anhedonic after exposure to stress, it is expected to drink less sugary solution than a rat that has not experienced stress. When tested for ICSS, anhedonic rats are expected to need stronger
stimulation to start self-stimulating. However, researchers have found that these changes are not always observed, and sometimes opposite responses are reported. This is a significant concern because the model is often used to screen for new antidepressant drugs and therapies. Using a model that does not produce consistent results can lead to misleading conclusions that risk developing drugs and therapies that are ineffective, wasting resources and animal lives in the process.
A limitation of the sucrose preference test is that water and food are frequently removed for an extended period before the test. This is to encourage rats to drink considerable amounts in a short period, which
makes the test more convenient for researchers. However, this practice causes the rat to be hungry and thirsty, which adds additional motivation to consume a caloric solution. This is also problematic in the
model because stressed rats often lose weight and have other metabolic changes as a consequence of stress. In this situation, it is hard to determine what the test measures because we do not know if the rat
drinks the solution because it likes it (it is pleasurable) or because of its metabolic needs. Our aims with the present thesis were to determine whether stressed rats drink less sugary solution when long periods of fasting are not used and whether more reliable results can be obtained with ICSS, a test that does not use sugar as a reward and it is therefore not influenced by factors that modulate the consumption of
food.

To answer these questions, we looked at published studies. Systematic reviews are a type of study that allows researchers to critically assess and synthesize all available evidence on a specific topic. This
comprehensive analysis helps to find biases and gaps in the current knowledge, which can guide researchers in designing more robust and reproducible experiments. Systematic reviews also significantly
contribute to a more humane use of laboratory animals. By reusing available data to reconcile conflicting results, they generate new data without using animals. They can also prevent redundant use, as the research question might have already been answered indirectly by published studies. Additionally, they can help to identify factors related to the model or the test that require fine-tuning, which can guide
refinement efforts in future research.

With our systematic reviews, we found that stressed rats overall consume less sugary solution than unstressed peers do. This demonstrates that fasting before the sucrose preference test is not necessary. Furthermore, we found that the difference in sweet consumption between stressed and unstressed rats was significantly smaller when a non-caloric sweetener was used instead of sucrose and the consumption
was adjusted for body weight. Additionally, longer periods of fasting made this difference appear larger than it is. This suggests that even in the absence of fasting, there are additional factors (possibly metabolic)
that confound the interpretation of the results. Thus, starving the animals is not only unnecessary but also unadvisable. We concluded that without properly controlling for additional factors that modify the
consumption of sugary solutions, the test is not a reliable measure of anhedonia. Yet, researchers have used this measure to distinguish animals that are susceptible to developing depression after stress from those that are not. When we dug deeper into the rationale behind this practice, we found no proper justification. What is more, we observed that researchers employing this practice analyse their data in ways that increase the risk of finding erroneous results that exacerbate the problem of unreliable results further. When we reviewed the evidence regarding the effect of the model on self-stimulation, we did not find support for the claim that stressed rats require stronger stimulation. Although pioneering studies in the area have suggested this, their results have not been replicated by others. Unfortunately, the evidence was so limited that we could not identify factors that may explain why results failed to be replicated.

Considering that we could not find evidence that the chronic unpredictable stress model leads to consistent effects on sweet consumption or self-stimulating behaviour, our results raise the question: is the model actually valid for the study of anhedonia and depression? It is crucial to continue the critical appraisal of the evidence supporting the validity of the model. By doing so, researchers can immensely benefit the research in the area and prevent subjecting animals to unnecessary harm.