It's very difficult to distinguish immobility and freezing behaviour, particularly using automated software. If you are scoring manually in a fear conditioning protocol, it is possible to look for the hunched posture, exaggerated respiration etc to identify freezing.

One other consideration is the length of time you are attempting to identify freezing for. If using rodents, the animals' natural inclination to explore an environment means that a lack of movement in a novel arena is usually due to 1 of 3 things:

1) Sedation/Muscle Relaxation effect (of a drug for instance)

2) Habituation to the arena

3) Fear-induced freezing

You should know whether a drug you have administered is sedative or muscle relaxant, so this can be ruled out as causing a lack of movement. Habituation occurs over a long (1-2 hours) period of time in rodents when the arena is completely novel, so a lack of movement in the first 5 minutes for example would not indicate habituation. Instead, a lack of movement within the first 5 minutes of placing in an arena in a fear conditioning experiment is more likely to suggest fear-induced freezing. Conversely, if immobility is occurring after a long period of time in an arena (1hr+), this is unlikely to be fear-induced freezing, and more likely habituation to the environment.

First tnx for reply. Yes, EthoVision software we have does not distinguish freezing from immobility and also in communication with other researchers most of them is using immobility as a measure of freezing behavior, which I am 100% convinced is not adequate measure. I cannot reconcile with that kind of thinking, I think its wrong.

Do you measure freezing behavior and how you do it? Manually or by software?

Also you made a very important statement, distinction between habituation to arena and fear-induced freezing and I agree with your thinking that in fear conditioning experiments a 3 min period reflects freezing and not just a habituation.

When I have measured freezing in the past I have done in manually, one animal at a time, using a stopwatch. I haven't tried to automate freezing behaviour, but as you say, I doubt the software would be sensitive enough to differentiate between the two. But if you do automate, I would expect the immobility measured by the software during the first 3-5 minutes to be almost solely freezing.

That is the point really it doesn't.  In the 3 min extinction period in mice we observe distinct differences between freeze and immobility. Visible to the naked eye, its so obviously frankly. I have some protocol to measure with stop watch every 5 or 10 seconds during 3 min period. Every 10th second you look to display and check for freeze + or -.

My answer may be naive.

Is freezing an active fear reaction and immobility just an absence of movement such as resting? I don't know the conditions inducing freezing in your animals, but

1- freezing could be in relation with some stimulus that you should monitor, so there is a close timely relationship between the two events

2- freezing is rapidly initiated, such as after fleeing. Therefore there may be a way to objectivize an abrupt change of activity of freezing onset vs a progressive slowing down if the animal enters resting.

3- A zooming up on sensory organs, such as e.g. whiskers movements, eyes or ears may give hints

4 Try to separate the two situations manually and look up on the automatized recording data if they are related. "Visible to the naked eye, its so obviously frankly" so try to explain someone how (writing it down), and if you succeed, then try teaching it to the machine.

Hopes this helps

Christian