Dr. Mogil’s Ph.D. dissertation, on the topic of stress-induced analgesia (SIA), sensitized MOGILab to the fact that stress can powerfully affect pain, and in both directions, via SIA or the equally robust stress-induced hyperalgesia (SIH). Although SIA and SIH are usually studied using stressors such as forced swimming or restraint, the laboratory environment (both husbandry and testing) is replete with potential stressors that might confound results. Over the years, MOGILab has described a number of previously unappreciated sources of lab environment stress.


In 1995, two groups identified the endogenous ligand of the “orphan” opioid receptor, ORL-1; the ligand was variously named “orphanin FQ” (Reinscheid et al., Science, 1995) and “nociceptin” (Meunier et al., Nature, 1995). Both groups reported that when injected into the brain (i.e., intracerebroventricularly, or i.c.v.) the ligand produced hyperalgesia. In a series of studies published between 1996 and 1998 with Dr. Judith Grisel, Dr. Mogil showed that both groups had failed to account for stress-induced analgesia (SIA) caused by the injection itself. In fact, i.c.v. orphanin FQ/nociceptin is functionally anti-opioid (although intrathecal orphanin FQ/nociceptin produces analgesia), and the apparent hyperalgesia was actually a blockade of the injection SIA.


Restraint is well known as a stressor, and restraining mice in Plexiglas can artificially inflate apparent heat pain withdrawal thresholds (i.e., cause SIA) compared to gentler restraint in a cloth/cardboard pocket, which MOGILab has long recommended for use in algesiometric testing in the mouse.

Experimenter Effects

An analysis of a large archive (>8000 measurements) of baseline tail-withdrawal test data collected in three different laboratories over 8 years revealed that the largest source of variability in the data was…the experimenter. A subsequent experiment designed to partition the variance suggested that the identity of the experimenter was responsible for almost 40% of the observed variability, compared to 27% for genotype. This might not seem so surprising, since in the tail-withdrawal test the experimenter is holding (via cloth/cardboard pockets; see above) the mouse being tested. In a more recent study, however, we observed that the mere presence in the room of male (but not female) experimenters can produce stress, and thus SIA (and thus artificially inflate apparent “baselines”), even in assays with no direct experimenter involvement during (video-based) data collection. This experimenter effect is olfactorily mediated, and could be recapitulated using chemosignals like androstenone, androstadienone, and especially, 3-methyl-2-hexanoic acid. MOGILab believes that this phenomenon may have confounded a large number of existing experiments.

Behavioural State

A critical but mostly ignored issue in algesiometry is precisely what the mouse or rat is doing at the precise time of testing. MOGILab categorized the behavioural state of non-locomoting mice (deep sleep, light sleep, resting, alert, or grooming) as heat or mechanical stimuli were applied to their hind paws. We found that grooming mice are robustly analgesic compared to all other behavioural states. This may be a mouse model of distraction analgesia.

Conditioned Hyperalgesia

Many experimental plans involve testing on multiple days. In an attempt to study the increasingly important topic of “pain memory”, MOGILab developed a new paradigm, in both mice and humans, of context-dependent, classically conditioned pain hypersensitivity using a single conditioning trial. We found that in both species, only males display such conditioned hyperalgesia, due to high stress levels on the second day of testing in males but not females. Beyond representing a nice mouse-to-human translation, this finding shows that stress can produced hyperalgesia (i.e., stress-induced hyperalgesia) instead of analgesia depending on the stress parameters. Under circumstances of repeated testing, male subjects will not exhibit stable “baselines”.

Pregnancy Chemosignaling

In a recently published paper, we have found that male (but not female) mice respond with stress and SIA to the proximity of late pregnant or lactating female mice. This effect is also mediated olfactorily, with the most effective chemosignal responsible being n-pentyl acetate. Amusingly, n-pentyl acetate is also the main odorant of bananas, and banana oil also produces stress in male mice.


This review paper summarizes the field of lab environment confounds.