Elderly populations face greater risks of mortality when exposed to changes in environmental stress. The purpose of the following study was to develop an age-dependent susceptibility model that achieved the following three goals: 1) to operationally define homeostasis by assessing the stability and periodicity in physical activity, heart rate (HR), and deep body temperature (T(db)), 2) to specify alterations in activity, HR, and T(db) regulation that signal imminent death, and 3) to test the hypothesis that the decay in homeostasis associated with imminent death incorporates the coincident disintegration of multiple physiological systems. To achieve these goals, the circadian regulation of activity, HR, and T(db) was assessed using radiotelemeters implanted in AKR/J (n = 17) inbred mice at approximately 190 days of age. During a 12:12-h light-dark cycle, weekly measurements were obtained at 30-min intervals for 48-h periods until each animal's natural death. The average (+/-SE) life span of surgically treated animals did not differ from untreated controls (319 +/- 12 vs. 319 +/- 14 days). Cardiac and thermal stability were characterized by a circadian periodicity, which oscillated around stable daily averages of 640 +/- 14 beats/min in HR and 36.6 +/- 0.1 degrees C in T(db). Stable HR and T(db) responses were compared with extreme conditions 3 days before death, during which a disintegration of circadian periodicity was coincident with a fall in the daily average HR and T(db) of approximately 29 and approximately 13% lower (i.e., 456 +/- 22 beats/min and 31.7 +/- 0.6 degrees C), respectively. The results further suggested that multiple predictors of cardiac and thermal instability in AK mice, including significant bradycardia, hypothermia, and a loss of circadian periodicity, forecast life span 5-6 wk before expiration.