Disease Models

The Basics of Depression Testing in Rodents

Major Depressive Disorder is one of the most common mental health diagnoses worldwide. Despite its astounding frequency, research has yet to uncover universally effective treatments for depression, let alone a cure for this chronic illness. Although, there are research which suggest possible natural treatments. Given its prevalence and the substantial hindrance major depressive disorder is to the daily lives of sufferers, there is much incentive to focus more resources on experiments through depression testing, for further comprehension of the neurological mechanisms that underlie depressive episodes.

As with any human psychiatric condition, modeling depression in animals for experimental purposes can be tricky. Many questions on depression testing need to be answered before we can draw any conclusions from our animal models of depression to humans. If we focus on rodent models (which we will for the purposes of this post), some of these questions include: What form does depression take in a rodent? How can we reliably and ethically induce a depressive phenotype in a rodent? What is the relative validity of different rodent depression assays? These important questions lay the foundation for depression testing in the lab, and we’ll address each of them right here.

What form does depression take in a rodent?

Behavioral testing in rodents generally relies on the identification of behaviors that can be considered analogous to the symptoms of the human condition. This is why papers so often include phrases like “the mouse displays a depressive-like phenotype” rather than “the mouse is depressed”. This distinction may appear inconsequential at first, but it is important to get into the habit of acknowledging that the phenotypes displayed by experimental animals are at best considered analogous to the human experience. When it comes to behavioral experiments like depression testing, we cannot be completely certain about what an animal is experiencing on any meaningful level with regard to cognition or mood, if for no other than the obvious reason that we cannot ask a rodent how it is feeling. Such a distinction also helps us to remember that the mechanisms underlying a mouse behavior, no matter how similar that behavior looks to a human symptom, is not necessarily the mechanism underlying a human behavior. These mechanisms can provide valuable insight into the human condition, but cannot be translated immediately between species at face value.

So what are these rodent behavioral analogs for depression symptoms? Two of the most common analogs are related to the human symptoms of anhedonia and despair.

Anhedonia refers to a loss of pleasure in things that a person once found enjoyable. In rodents, this is generally reflected by a decrease in the preference for known rewards, such as sucrose.

Despair refers to a loss of hope, generally accompanied by feelings of sadness. In rodents, this is modeled by placing the animal in an aversive situation and measuring the extent to which the animal tries to remove itself from that environment.

These are very useful analogs for depression testing because they are related to natural rodent behaviors, meaning that a healthy rodent will naturally try to obtain rewards or escape aversive stimuli regardless of any prior experimental manipulation, so a reluctance to demonstrate either of those behaviors can reasonably be considered abnormal. We’ll discuss how to scientifically measure these behaviors when we talk about depression assays a little later on.

Depression Testing: How to induce depressive-like phenotype in rodents

Depression Testing: How to induce depressive-like phenotype in rodents

It is generally accepted that depression is the result of both genetic and environmental factors. As such, it is possible to evoke a depressive-like phenotype with either a genetic model or with environmental manipulation for depression testing.

Environmental stressors are widely used to evoke a depressive-like phenotype in rodents. Before beginning any such protocol in depression testing, it is important to review all guidelines thoroughly to ensure that you are performing your experiments in the most ethical way possible. By definition, these protocols will be stressful to the animal subjects. It is your responsibility to ensure that no undue stress is caused to your animals during your work. This is critical not only to comply with ethical standards in research, but also to ensure the validity of your findings later on.

In depression testing, one very common environmental stressor used to simulate adverse early life events is maternal separation. This entails the premature separation of pups from their mother. This procedure occurs before the weaning age (around P21) and entails separating mother and pups into different cages for a few hours per day. Often, this kind of early life stress is used to create an animal that is more vulnerable to developing a depressive-like phenotype during depression testing, when stressed as an adult than a typically raised animal.

Altering the home environment of the mouse can modulate its basal stress levels.[3] While such manipulations may not induce a depressive-like phenotype per se, a more stressed mouse may be more vulnerable to developing depressive behaviors later on. Various aspects of the home cage environment, including the presence of littermates, enrichment, or access to exercise, can influence the general stress level of the animal. A mouse that lives alone in a small, empty cage may perform differently in a depression assay than a mouse that lives with two of its littermates in a spacious cage with a running wheel and other enrichment. Care should be taken to ensure that the mice in a given behavioral experiment cohort have similar living environments to minimize potential confounds.

During depression testing in adult animals, depressive-like phenotypes can be induced with either chronic or acute stress. Generally speaking, chronic stress tends to be the preferred protocol over acute stress.

Acute stress entails a single exposure to severe stress. This can be difficult to achieve in a laboratory setting, as we do not know the subjective experiences of a “severe” versus a “mild” stressor for a mouse. Individual animals (just like us humans) vary widely in their susceptibility to depression, and it can be difficult to develop an ethical, universally severe acute stressor that will produce reliable results. In addition to being experimentally tricky, human depression is often considered to be a chronic condition that develops over time anyway, so chronic stress may be more relevant to major depressive disorder research.

Chronic stress is generally administered over the course of several weeks for rodents, and in events that are more mild than would be needed for acute stress models. One example of this is restraint stress, in which an animal’s mobility is temporarily constrained, administered periodically over a prescribed amount of time. Schedules and intensities of stressors vary between studies and require specific consideration for the goals of each experiment.

On the genetic side, new mutant mouse models for depression testing are regularly developed. Many of these models rely on knocking out a known candidate gene for depression, and observing the effects. A comprehensive review of genetic mouse models of depression is beyond the scope of this post. If you want to work with a genetic model and don’t know which model is the best choice for you, a good starting resource can be found in [1].

Best ways to measure a depressive-like phenotype

There are several frequently used behavioral assays that are used in depression testing. Before beginning a project, it is important to consider the relative strengths and weaknesses of these assays to determine which is most appropriate for the question you want to answer. For example, some tests are sensitive to the effects of antidepressants, some are particularly insensitive, and others still may reflect chronic but not acute antidepressant treatment. Assays will also vary in their degrees of experimental validity. When considering validity in depression testing, there are three basic domains: face validity (i.e. does the mouse behavior look like the human behavior), construct validity (i.e. is the mechanism in question the same as that which underlies the human behavior), and predictive validity (i.e. can the response of the mouse predict the response of the human).[2]

Here we will review the most popular depression assays, as well as their relative strengths and weaknesses in depression testing. Often these tests are used in some combination with one another, and sometimes they are part of a larger battery of behavioral assays. If this is the case, it is important to consider the relative stress level of each test. For example, a forced swim test is considerably more stressful to a mouse than an open field test. These more stressful tests should generally be performed last, after less stressful tests have been completed. Failure to organize behavioral assay batteries carefully can produce skewed results.

Best ways to measure a depressive-like phenotype depression testing in rodents

Forced Swim Test

The forced swim test is one of the popular depression assays used in depression testing.

Description: A mouse is placed into a cylinder of room temperature water. After two minutes of habituation, immobility is scored for four minutes.

Interpretation: A mouse that spends more time immobile than swimming is considered to be modeling a state of “behavioral despair.”

Face Validity: Okay. Immobile mice do look like they have given up, though this interpretation is the result of substantial anthropomorphizing.

Construct Validity: Not great. An immobile mouse could just be saving energy, which is actually a great strategy to keep its head above water.

Predictive Validity: Pretty good. Antidepressants generally reduce time spent immobile, and do so better than other kinds of drugs. Some antidepressants are better than others.

To learn more about this test, visit the Forced Swim Test portfolio page.

Novelty-Suppressed Feeding

Description: After a period of food deprivation, a mouse is placed in a novel environment and a food pellet is placed in the center of the testing arena. Latency to begin eating is measured, generally with a maximum time limit decided beforehand.

Interpretation: Anxious mice will take longer to begin eating.

Face Validity: Pretty good. Chronic mild stress increases the latency to approach the food.

Construct Validity: Good. Response to antidepressants takes time, just as in humans. Decreased latency following antidepressant treatment is also associated with hippocampal neurogenesis, which is considered beneficial for humans suffering from depression.

Predictive Validity: Good. Antidepressant treatment – but only chronic – decreases response latency.

Sucrose Preference Test

Description: Mice have the option to drink regular water or water that contains sucrose. Consumption of each is compared.

Interpretation: Mice should prefer the sweet sucrose water, and mice that do not may be considered to be exhibiting anhedonia.

Face Validity: Great. Similar tests in humans yield similar results, and antidepressants only rescue the response if the mouse has been exposed to stress previously.

Construct Validity: Pretty good. There is likely more overlap in mechanism for sucrose preference between species than there is for time spent immobile in other tests.

Predictive Validity: Pretty good. Antidepressants rescue sucrose preference, but only if the mouse has been exposed to stress previously.

Tail Suspension Test

Description: A mouse is suspended upside down by the tail. After two minutes of habituation, immobility is scored for four minutes.

Interpretation: A mouse that spends more time immobile than attempting to right itself is considered to be modeling a state of “behavioral despair.”

Face Validity: Okay. Like the FST, we’re anthropomorphizing to interpret the immobility.

Construct Validity: Poor.

Predictive Validity: Pretty good. Antidepressants generally reduce time spent immobile. Some antidepressants work better than others.

Visit our Tail Suspension portfolio page to understand more about the test.

There are many ways to test for a depressive-like phenotype when it comes to depression testing in mice. It is common practice to use a few of these assays in the same project, but don’t forget that the order of testing matters! You don’t want to conduct a stressful test early in your project that will have unintended impacts on subsequent testing. Finally, as with any behavioral experiment, when carrying out a depression test in rodents, pay close attention to the types of measures that you need to be collecting, and especially to the valid ways in which that data can be interpreted. Not all depressive behavior assays are created equal, and certain assays will be better suited to answer your questions than others.


  1. Cowen, PJ, Sharp, T, and Lau, JYF (Eds). Behavioral Neurobiology of Depression and Its Treatment, Current Topics in Behavioral Neuroscience Vol. 14.
  2. Powell, TR, Fernandes, C, and Schalkwyk, LC. (2012). Depression-related behavioral tests. Current Protocols in Mouse Biology 2: 119-127.
  3. Gurfein, BT, Stamm, AW, Bacchetti, P, Dallman, MF, Nadkarni, NA, Milush, JM, Touma, C, Palme, R, Di Borgo, CP, Fromentin, G, Lown-Hecht, R, Konsman, JP, Acree, M, Premenko-Lanier, M, Darcel, N, Hecht, FM, and Nixon, DF. (2012). The calm mouse: an animal model of stress reduction. Mol Med 18: 606-617.

About Caroline Sferrazza

Caroline Sferrazza is a graduate student in neuroscience at University of California, San Diego. She is working on her thesis in Dr. Lawrence S.B. Goldstein’s laboratory at the Sanford Consortium for Regenerative Medicine, where she uses human induced pluripotent stem cells to model and study Alzheimer’s Disease. As a science communicator, Caroline is a regular contributor to UCSD’s NeuWrite blog, as well as other online scientific resources, and is an active organizer of local science outreach events in San Diego, including school visits and science festivals like Taste of Science and ComSciCon.