receptors in the adrenal cortex, resulting in the release of glucocorticoids (cortisol or corticosterone), which mobilize energy stores in response to the perceived stress. When the stressor is removed or otherwise adapted to, glucocorticoids bound to receptors in the hypothalamus and pituitary initiate negative feedback that causes a decrease in the production and release of CRH and ACTH, thus terminating the hormonal response and completing the negative feedback loop (Meaney et al. 1996; Miller and O’Callaghan 2002). It should be noted that the HPA axis affects many hormonal and neural systems and plays a role in modulating the immune system.

Both positive and negative stimuli activate the HPA axis with short-and long-lasting effects. For example, exposure to novel stimuli may elicit exploratory behavior and brief activation of the HPA axis. In contrast, prolonged or repeated stressors, such as social separation of young from their mother, generally elicit strong protest reactions and activation of the HPA axis (Levine 2005; Levine and Mody 2003; Vogt et al. 1980). In the first instance, homeostasis is quickly restored, whereas in the latter case animals may be subjected to chronic HPA changes associated with neuroendocrine stress resistance that persists even after animals are returned to their mother (Parker et al. 2006).

Glucocorticoid levels (usually corticosterone in rodents, cortisol in other species) are used as indicators of the strength and impact of a stressor. Meaningful interpretation of these values, however, presents significant challenges. Glucocorticoids are typically measured in blood serum or plasma but can also be quantified in saliva, urine, feces, and hair (Abelson et al. 2005; for more references see Additional References).

Blood sampling requires venipuncture and possibly other stressful procedures such as handling, transport, capture, restraint, needle stick(s), and sedation. Unless animals are habituated to blood sampling, the method itself can activate the HPA axis thereby confounding assay results. A less stressful sampling method involves measuring glucocorticoid levels in hair. Hair samples are obtained by shaving hair from a particular region (usually the nape of the neck) and then shaving the hair once again after a defined period of regrowth (for a discussion of the method in primates see Davenport et al. 2006). Although this procedure requires the animals’ habituation, restraint, or sedation, unlike venipuncture the stress caused by hair collection does not confound the measurement. Similarly, saliva collection may impact animals less if they have been habituated to the process (Lutz et al. 2000). Urine and feces are collected after excretion from the body and so probably have the least impact, unless animals are not habituated to the special metabolic cages used for sample collection.

The type of sample obtained and the time frame it reflects may also influence results. Blood and saliva yield an index of stress at one brief moment in time (point samples) and are, therefore, influenced by circadian



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement