Life in the Cold: Considering Laboratory Animal Housing Temperature as an Important Variable in Shock and Trauma Research.

Preclinical research relies heavily on appropriate animal models to better understand human biology and to test the safety and efficacy of new therapies. It is of upmost importance that researchers continually strive to optimize preclinical models, both from the standpoint of animal welfare, and to promote the translatability of preclinical data to the clinic.

Guidelines in the US recommend housing laboratory rodents at temperatures ranging from 20-26°C (1). Most vivaria that I have worked in seem split the difference, housing rodents at around 22-24°C. I have always assumed that this temperature range has become the norm as it is one that is comfortable for PPE clad humans – something I can attest to having recently had building maintenance at my institution adjust the ambient temperature in one of our rodent holding rooms to 30°C. However, there is an abundance of empirical data indicating that housing temperatures in the low to mid 20°Cs likely represents mild cold stress for rodents (see (2) for a detailed review on rodent thermoneutrality).

Rodents can maintain core temperature over a wide range of housing temperatures (3). Housing rodents at temperatures close to freezing have long been used to study behavioral and metabolic adaptations to cold exposure. More recently, the metabolic impact of a life time being a little cold – that is housing rodents a few degrees centigrade below thermoneutrality over their lifetime – has come under scrutiny. This has been driven in part by translational failure of certain therapies proven effective in rodents. Shock and trauma research has not been immune to scrutiny with the regards to the utility of preclinical murine models (4). Having been involved in research focused on the stress response to burn trauma, I often wonder what impact housing temperature might have with regard to translational value of my own rodent data. Burn trauma is perhaps a good example to illustrate the potential impact of rodent housing temperature in terms of accurately recapitulating response seen in patients, since housing patients in warmer temperature reduces heat loss and thus hypermetabolism (5, 6).

There is a clear difference in total energy expenditure (TEE) in rodents housed at 30°C compared to those housed at 20°C or lower (2, 3). However, since most laboratory rodents reside at temperatures (22-24°C) that are perhaps around 5°C below thermoneutrality, I was curious as to whether more subtle changes in housing temperature had on TEE. So - I decided to do a little experiment. I put a bunch of C57BL6 mice in metabolic cages for about a week. As a side note - the metabolic cages I have access to are pretty neat. They perform indirect calorimetry in the pull mode, where air is pulled out of the animals’ cage at a pretty high flow rate. This approach has a couple of advantages. First, the cage is not pressurized and closely resembles a rodent home cage (complete with cage bedding). Second, high flow rates provide high resolution respiratory gas exchange data, allowing energy expenditure data can be determined from each cage every few minutes (rather than 2-3 times an hour like some older rodent metabolic cage systems). This means that in addition to determining TEE, you can also calculate various components of TEE such as resting energy expenditure (REE).

I am fortunate that the metabolic cage system I have access to is housed within environmental cabinets, which make it really easy to change the animals housing temperature. So - after a few days of hanging out at ~24°C (the normal temperature in our animal vivarium), I increased the temperature within the environmental cabinets the mice were house in to ~30°C. Just a 6°C difference. The TEE and REE data from this little experiment are presented in Figure 1.  While I fully expected that transitioning mice from 24°C to 30°C would reduce energy expenditure – I was a little surprised that a 6°C increase in temperature resulted in a ~20% and ~30% reduction in TEE and REE, respectively. Assuming that 30°C is pretty close to thermoneutrality for a mouse – you could look at these data a different way. Taking TEE and REE at 30°C as ‘normal baseline’ values – housing these mice at 24°C results in a 45% increase in REE (and 27% increase in TEE). Accordingly – it’s not too much of a stretch to suggest that certain ambient temperatures - which we might think represent ‘normal’ for housing rodents - may in fact induce rather marked hypermetabolism.


Figure 1. Total (TEE) and resting (REE) expenditure in male (n=16) mice housed at 24°C before being transitioned to 30°C (Porter et. al., unpublished observation). TEE and REE data where determined by indirect calorimetry in a Promethion Core system (Sable Systems, Las Vegas, NV). Values for 24°C and 30°C are the average of 2-3 consecutive days.


While ambient temperature can clearly have a significant impact of REE and TEE in rodents, I should note, there are many other factors that can influence rodent energy expenditure. These include housing density, the type and amount of bedding material provided, cage design, and whether running wheels are provided or not. And of course - different mouse strains may have different thermoneutral zones (2). With that said, since ambient temperature is a variable that can be controlled relatively easily – and given that critically ill patients and those with significant trauma are often transiently hypermetabolic - a reasonable question many of us that use rodents to study responses to trauma and critical illness might ask is – does housing temperature impact the translatability of my preclinical rodent data?

Some pretty interesting data published in SHOCK last year provided a few answers to this question. Dr. Carpenter and colleagues hypothesized that housing mice at thermoneutrality (30°C) after a cecal ligation and puncture procedure would alter immune responses and improve survival when compared to mice housed at 22°C after cecal ligation and puncture (7). The authors found that 78% of mice housed at 30°C survived this model of peritonitis, whereas significantly fewer mice (only 40%) housed at 22°C survived. Note that a loss of a righting reflex was used as a humane endpoint for survival studies (7). These survival data were complemented with endpoint studies that showed core temperature was lower in mice housed at 22°C versus 30°C at 24 hr. post cecal ligation and puncture. Further, housing mice at 30°C markedly lowered bacterial colony forming units in peritoneal lavage fluids compared to mice housed at 22°C via an increase in local phagocytosis.

While it is unclear if there are optimal temperatures to house rodents in studies involving the induction of sepsis – or any other form of trauma or critical illness for that matter, the interesting experiments performed by Carpenter and colleagues nicely demonstrate the impact of housing temperature on bacteremia and survival in a mouse model polymicrobial peritonitis. Clearly then, housing temperature is likely an important variable to consider when establishing models and designing experiments – particularly in the context of critical illness and trauma.

While I realize this blog post poses plenty of questions while offering little in the way of concrete answers – the growing emphasis placed on rigor and reproducibility in biomedical research coupled with questions relating to utility of some rodent models to accurately recapitulate complex human diseases, researchers must constantly challenge themselves to improve and refine preclinical models. In my opinion, the temperatures we house our laboratory rodents at is an important and readily modifiable parameter that may help improve preclinical models and ultimately promote the translation of basic science to the clinic.


About the Blogger:

Craig Porter is an Associate Professor at the University of Arkansas for Medical Sciences. His interests include understanding the role of the mitochondrion in the hypermetabolic response to burns. Craig joined the Shock Society as a postdoctoral fellow in 2014. He currently serves on the societies’ communications committee.



1.         Institute for Laboratory Animal Research. Guide for the Care and Use of Laboratory Animals. The National Academies Press. 2011;Eight Edition.

2.         Gordon C. Thermal physiology of laboratory mice: Defining thermoneutrality. J Therm Biol. 2012;37:654-85.

3.         Speakman J. Measuring energy metabolism in the mouse - theoretical, practical, and analytical considerations. Front Physiol. 2013;4:4:34.

4.         Seok J, Warren H, Cuenca A, Mindrinos M, Baker H, Xu W, Richards D, McDonald-Smith G, Gao H, Hennessy L, Finnerty C, López C, Honari S, Moore E, Minei J, Cuschieri J, Bankey PJ, JL, Sperry J, Nathens A, Billiar T, West MJ, MG, Klein M, Gamelli R, Gibran N, Brownstein B, Miller-Graziano C, Calvano S, Mason P, Cobb JR, LG, Lowry S, Maier R, Moldawer L, Herndon D, Davis R, Xiao W, Tompkins  R, Inflammation and Host Response to Injury LSCRP. Genomic responses in mouse models poorly mimic human inflammatory diseases. Proc Natl Acad Sci U S A. 2013;110(9):3507-12.

5.         Wilmore D, Long J, Mason AJ, Skreen R, Pruitt BJ. Catecholamines: mediator of the hypermetabolic response to thermal injury. Ann Surg. 1974;180:653-69.

6.         Caldwell FJ, Wallace B, Cone J, Manuel L. Control of the hypermetabolic response to burn injury using environmental factors. Ann Surg. 1992;215:485-90; PMCID: PMC1242481.

7.         Carpenter K, Zhou Y, Hakenjos J, Fry C, Nemzek J. Thermoneutral Housing Temperature Improves Survival in a Murine Model of Polymicrobial Peritonitis. Shock. 2020;54(5):688-96.


Share this post:

Comments on "Life in the Cold: Considering Laboratory Animal Housing Temperature as an Important Variable in Shock and Trauma Research. "

Comments 0-5 of 1

Melanie Scott - Tuesday, August 03, 2021

Really thought-provoking, Craig, especially as so many of us are starting to think more about metabolic processes and how they impact inflammation and immunity.

Please login to comment