Body condition scoring: a rapid and accurate method for assessing health status in mice.

Practical, rapid, noninvasive methods for assessing healthstatus and establishing endpoints are needed in mouse ex-periments in which wasting and death are a potential end-point, including aging and toxicology studies, ascitesproduction, and phenotype analysis in mutant mouse colo-nies. Current methods for assessing the health status of amouse and establishing endpoints might include observa-tion of behavior, assessment of physical appearance, andmeasurement of body weight (BW). Behavioral parametersinclude observation of unprovoked behavior and responsesto external stimuli. Classic changes in physical appearanceinclude exophthalmia or enophthalmia (bulging or sunkeneyes, respectively), nasal or ocular discharge, rough coat, andhunched back. These observations, as well as additional onesparticular to an experimental procedure or to the geneticmakeup of an animal, have been suggested as standard indi-cators of ill health (1–4). These clinical indicators can bescored as degree-of-deviation-from-normal, thereby allowingan animal to be monitored over time as health declines (1–4). Decreased food and water consumption is an importantsign of deteriorating health (4), which generally results inloss of BW; support for using weight loss as an indicator ofpoor health in rodents comes from the study by Redgate etal. (5). They determined that 7 or more consecutive days ofweight loss in central nervous system tumor-bearing ratscorrelated well with irreversible progression to death. How-ever, Beynen et al. (2) found that observation of behavior andphysical appearance was largely ineffective for discriminat-ing between gallstone-bearing mice and healthy controls,and weight loss was significantly different between the twogroups of mice for the males but not the females. They con-cluded that response to palpation of the right hypochon-drium (i.e., signs of a painful response) was the bestindicator of gallstones for males and females.Twenty percent loss of rodent BW or prolonged weightloss (progressing to an emaciated state) are generally estab-lished criteria for euthanasia (3, 6). However, there are prac-tical problems with the measurement of BW, which may notyield an accurate measure of fat stores and muscle mass,because reduction of fat stores and muscle mass (as mea-sured by BW) is masked if weight loss is displaced by tumorgrowth, organ enlargement, or intraperitoneal fluid accumu-lation. Furthermore, the reference weight of a healthymouse will vary according to sex, age, body frame size, andin females, pregnancy status. Scoring body condition (BC) byobserving the amount of flesh covering bony protuberancesis largely independent of the aforementioned confoundingvariables. The technique of BC scoring as a method forevaluating animal condition and nutritional state has al-ready been validated for use in dairy cows (7–9), beef cows(10), goats (11), sheep (12), and horses (13). In cows, BC scorecorrelates with the amount of subcutaneous fat stores (7, 10)and nutritional status (14). Additionally, abnormal loss of BCwas found to be an indicator of mastitis in dairy cows (15).Our goals were to evaluate the accuracy of the BC scoringtechnique in assessing the health of mice that have organ en-largement concurrent with declining health; compare the accu-racy of this method with that of using BW for assessing healthstatus in these animals; and determine the interobserver reli-ability of the BC scoring technique. P- and E-Selectin doubledeficient (P/E