PHYSIOLOGICAL REVIEWS   Vol. 78 No. 2 April 1998, pp. 339-358
Copyright ©1998 by the American Physiological Society

Cold-Induced Thermoregulation and Biological Aging

MARIA FLOREZ-DUQUET AND ROGER B. McDONALD

Physiology Graduate Group and Department of Nutrition, University of California, Davis, California

I. INTRODUCTION
    A. Definition of Aging
    B. Designs in Research on Aging
II. OVERVIEW OF THE EFFECTS OF AGING ON COLD-INDUCED THERMOREGULATION
    A. Human Investigation
    B. Animal Research
    C. Cold Perception
III. HEAT LOSS
    A. Vasoconstriction, Heat Loss, and Aging
    B. Insulative Properties of Body Fat and Aging
IV. HEAT PRODUCTION
    A. Skeletal Muscle Shivering Thermogenesis and Aging
    B. Brown Adipose Tissue Nonshivering Thermogenesis and Aging
V. COLD-INDUCED THERMOREGULATION AND THE RATE OF AGING
VI. SUMMARY AND CONCLUSIONS
    A. Summary
    B. Conclusions
REFERENCES

	ABSTRACT

Top References

Florez-Duquet, Maria, and Roger B. McDonald. Cold-Induced Thermoregulation and Biological Aging. Physiol. Rev. 78: 339-358, 1998.  Aging is associated with diminished cold-induced thermoregulation (CIT). The mechanisms accounting for this phenomenon have yet to be clearly elucidated but most likely reflect a combination of increased heat loss and decreased metabolic heat production. The inability of the aged subject to reduce heat loss during cold exposure is associated with diminished reactive tone of the cutaneous vasculature and, to a lesser degree, alterations in the insulative properties of body fat. Cold-induced metabolic heat production via skeletal muscle shivering thermogenesis and brown adipose tissue nonshivering thermogenesis appears to decline with age. Few investigations have directly linked diminished skeletal muscle shivering thermogenesis with the age-related reduction in cold-induced thermoregulatory capacity. Rather, age-related declines in skeletal muscle mass and metabolic activity are cited as evidence for decreased heat production via shivering. Reduced mass, GDP binding to brown fat mitochondria, and uncoupling protein (UCP) levels are cited as evidence for attenuated brown adipose tissue cold-induced nonshivering thermogenic capacity during aging. The age-related reduction in brown fat nonshivering thermogenic capacity most likely reflects altered cellular signal transduction rather than changes in neural and hormonal signaling. The discussion in this review focuses on how alterations in CIT during the life span may offer insight into possible mechanisms of biological aging. Although the preponderance of evidence presented here demonstrates that CIT declines with chronological time, the mechanism reflecting this attenuated function remains to be elucidated. The inability to draw definitive conclusions regarding biological aging and CIT reflects the lack of a clear definition of aging. It is unlikely that the mechanisms accounting for the decline in cold-induced thermoregulation during aging will be determined until biological aging is more precisely defined.

	I. INTRODUCTION

Top Next References

An inability to appropriately thermoregulate during cold exposure has been established in aged humans and in laboratory animals (22, 73, 79, 94, 107, 148, 162, 166, 181). The mechanisms accounting for this phenomenon have yet to be clearly elucidated but most likely reflect a combination of increased heat loss and decreased metabolic heat production. Although the current data strongly suggest that the decreased cold-induced thermoregulation reflects a biological age-related phenomenon, clinical hypothermia, a syndrome leading to death, in humans reflects more closely socioeconomic factors; that is, clinical hypothermia in the elderly is easily preventable by increasing the room temperatures and/or increasing insulation from clothing (i.e., putting on a coat). For this reason, we believe that the study of cold-induced thermoregulation is better suited as a model to evaluate possible mechanisms that reflect biological aging. Thus the discussion in this review focuses on how alterations in cold-induced thermoregulation (CIT) during the life span may offer insight into possible mechanisms of biological aging.

Research on the cold-induced temperature regulation during senescence is complicated by the fact that a precise and generally accepted definition of aging has yet to be adopted. Lack of a generally accepted definition of aging has hampered elucidation of possible mechanisms that reflect decreased CIT during aging; that is, comparisons of investigations among laboratories are often complicated by difference of age groupings, unknown health status of the population, a diversity of species being used as models, and lack of valid biomarkers that determine biological age. These complex "problems" surrounding the development of a definition of aging have often resulted in misinterpretation of data concerning CIT. It is important, therefore, that we begin this review with a brief discussion focusing on definitions of aging and experimental designs for aging research. This section is followed by an in-depth analysis of possible mechanism(s) that may account for age-related alterations in CIT. Finally, we present experimental evidence suggesting that the study of CIT can be used to evaluate mechanisms of biological aging at the individual level.

The fundamental biological mechanisms that underlie the aging process are unknown. As such, there has been little agreement among researchers as to a useful or precise definition of the word aging. Aging, as defined generally throughout the biological literature, is the gradual deterioration in function that occurs after maturity. This definition is imprecise, however, because biological aging includes all processes, both advantageous and detrimental, occurring within the organism from the beginning of life (conception/birth?) until death. Some researchers prefer to define biological aging more narrowly by using the word senescence. Senescence is generally accepted to mean a time-dependent loss in function that leads to disability or death. This definition of aging also lacks precision because it fails to include processes in early life or in utero that are recognized as having significant affects on adult function, disease, and life span (7, 8). In this review, the words aging and senescence are used interchangeably and, unless otherwise noted, refer more closely to the definition of senescence as used above.

Understanding the definition of aging is more important to the study of gerontology than simply a starting point for this review. The definition of biological aging is basic to the development of hypotheses-driven research. If one adheres to a broad definition of aging that includes all biological processes throughout the life span, then one must consider experimental designs that include the study of early as well as late life experiences (or how early events may affect later outcomes). Those choosing to investigate aging at the point where deterioration of function is clearly identifiable will focus on the upper end of the life span; that is, the results of any particular investigation on biological aging will be applicable only as far as the underlying definition of aging permits.

In our studies on aging, we have taken the position that aging can be defined as the inability of the organism to maintain homeostatic regulation when given a challenge. Any alteration in homeostasis must be nonreversible, independent of a pathological condition, and contribute significantly to a general loss in function or death. This definition is similar to that recently presented by Masoro (104), who suggests that aging represents "detrimental changes with time during postmaturational life that underlie an increasing vulnerability to challenges thereby decreasing the ability of the organism to survive." Vulnerability to challenges implies alterations in homeostatic regulation and, therefore, a decrease of integration among several physiological systems. To investigate biological aging from this viewpoint, it is advantageous to use a model system that includes many points of regulation. We have found that the evaluation of CIT provides an acceptable system in which to test hypotheses within the framework of our broad definition of aging.

The development of efficient experimental designs to test hypotheses concerning biological aging presents significant challenges that are rarely encountered in other biological research. Methodological issues for design and analysis of aging studies have been reviewed (27, 29, 41, 187) and are discussed only briefly. The primary purpose of aging research is to determine a biological change occurring as a function of time that is associated with the life span of the species. Cross-sectional designs using humans and rodents, by far the most common model, are capable of evaluating apparent time-dependent changes but are limited in their capacity to determine whether these alterations are associated with a basic process of biological aging. This is due, in part, to the fact that cross-sectional designs evaluate the biological process between young and old age groups at a single point in time. There can be no assurance that the measured variable at the given time point will remain constant in the future or will be representative of the entire population. Without repeated measurement over time, it is impossible to determine if the result reflects aging per se or an uncontrolled environmental factor or disease. Indeed, several early cross-sectional studies using humans to evaluate CIT included subjects recruited from a clinical populations. Moreover, cross-sectional designs in humans typically base their criteria for inclusion of individuals into the oldest grouping on an arbitrary definition, i.e., those over 65 years, rather than on more empirically derived data, i.e., survival or mortality distributions. This often results in the oldest age group having a wide distribution of ages. Significant biological differences undoubtedly exist among individuals at 65 vs. 90 yr of age just as they do between a 10 and a 20 yr old. Cross-sectional studies in humans are better suited for the development rather than the testing of hypotheses.

Cross-sectional studies of humans are not well suited for evaluating possible mechanisms of biological aging. Animal models, including rodents, insects, birds, and fish, offer the investigator an opportunity to study aging in a highly controlled environment. Important markers of biological aging such as longevity characteristics and late-life disease patterns can be determined easily in a variety of species. The investigator has significant control of environmental factors including diet, husbandry, and genetic characteristics. Nonetheless, there are several issues in the design of animal experimentation for the biology of aging that must be considered to appropriately interpret the results. Efficient cross-sectional design of animal experimentation requires a fairly precise determination of the age at which a species becomes senescent and, for reasons of comparison, the age that the animal reaches maturity. Correctly determining the latter has proven to be as important to proper design of aging research as the former. The results from several investigations suggest that developmentally immature animals (generally accepted as the age at which exponential growth rate is occurring) are not a suitable control group for senescent animals (103, 187). Rapidly growing versus weight-stable older animals have significantly different hormonal profiles that may affect various physiological systems. Conclusions based on cross-sectional comparisons between rapidly growing young and weight-stable old animals more likely reflect developmental changes rather those associated with senescence.

Determining the age at which a species "passes" into senescence is also difficult. The mean life span for a species as determined from survival curve analysis is used extensively as a marker of senescence. Mean values, including mean life span, are associated with a distribution of the individual data points, thereby resulting in considerable heterogeneity in the population value; this is particularly true in old animals. Age-related heterogeneity introduces significant error not often accounted for by standard error-comparison statistics. For example, the generally accepted mean life span of the Fischer 344 (F344) rat, a species used extensively in aging research, is 24-25 mo of age (192). The exponential decline in survivorship begins at ~20 mo of age and may continue until 29 mo (Fig. 1). This distribution of survival illustrates important issues that are rarely controlled for when using mean life span as a criterion for senescence. The cause of death in animals that die at 20 or 29 mo is rarely compared with the animals at the median life span of the group. It is possible that the cause of biological aging may differ significantly among these age cohorts. Differences in possible cause of death among aged animals and humans also illustrate the most serious problem of using cross-sectional design: selected mortality; that is, it is possible that the results of these investigations on a particular older age group more closely reflect phenomena associated with survival rather than biological aging. Moreover, median life span is based on chronological time. Although chronological time provides an easy marker for senescence in a group, it does not consider that the rate of aging is specific to the individual; that is, median life span of a species does not reflect biological aging.

View larger version (12K): [in this window] [in a new window]   FIG. 1.   Survival curves for male Fischer 344 rats fed ad libitum (group A, solid line) or fed a calorie-restricted diet (group R, dashed line). [From Yu et al. (192).]

	II. OVERVIEW OF THE EFFECTS OF AGING ON COLD-INDUCED THERMOREGULATION

Top Previous Next References

The physiological processes that allow maintenance of normothermia during cold exposure involve a complex series of regulatory steps and are described thoroughly elsewhere (55, 65, 77, 142). The high level of regulatory control of CIT has made the study of CIT an attractive and useful system in which to investigate possible mechanisms for aging. Briefly, upon exposure to a temperature below thermoneutrality (28-33°C in humans, Ref. 191; 24-30°C in rats, Refs. 55, 112), thermoreceptors in the skin, spine, and brain transmit neural signals to the hypothalamus. The integration of these neural signals results, in the case of cold exposure, in decreased heat loss and increased heat production. Heat is conserved during cold exposure by increasing cutaneous vasoconstriction, and thereby minimizing the heat exchange capacity of the skin. Concomitantly, heat production is increased as a result of shivering in skeletal muscle and nonshivering thermogenesis in a variety of peripheral tissues, but most notably in brown adipose tissue (BAT). In the event that the heat produced is insufficient and/or heat is not appropriately conserved during cold exposure, hypothermia ensues. Extreme hypothermia may lead to stroke or myocardial fibrillation and, eventually, death.

Initial reports of altered CIT in human elderly are based largely on observations in clinical settings. These descriptions suggest that hypothermia in the cold-exposed elderly reflects a loss of metabolically active tissue, poor circulation, and a slow adjustment to changing temperature, although no measurements of these variables were conducted (22-24, 88, 186). Concluding that CIT is reduced during aging based on investigations using elderly hypothermia victims may be inappropriate because the health status of these patients before cold exposure is rarely known; that is, disease rather than biological aging may influence greatly the observed hypothermia. Nonetheless, the susceptibility of the elderly to hypothermia is consistent with the finding that ~9% of elderly in the United Kingdom have morning core temperatures within one-half a degree of the clinical definition of hypothermia, 35°C (49). The contribution of biological aging to the low core temperature observed in these individuals is unknown because inadequate home heating and socioeconomic factors accounted for a significant portion of the variance. Moreover, elderly individuals, aged 69-90 yr, living in conditions of adequate home heating did not experience significant declines in core temperature or incidences of hypothermia (23). Although some clinical and epidemiological evidence may support the suggestion that CIT declines with aging, the influence of biological aging on the ability to maintain normal body temperature during cold exposure is difficult to precisely determine from these types of investigations.

To more accurately determine the effects that biological aging may have on CIT, investigations in humans should include subjects in which health status is clearly defined and the time of exposure and the ambient temperatures are controlled within narrow limits. Healthy older individuals exposed to cold (0-17°C) for periods ranging from 30 min to 2 h do not tolerate the exposure to the same degree as do younger subjects (Table 1; Refs. 42, 72, 73, 105, 181, 182). Rectal temperature of men aged 46-67 yr exposed to 17°C for 30 min is significantly lower than that observed in boys between the ages of 10 and 16 yr, but not compared with males 19-29 yr of age (182). Others have found that although core temperature does not differ significantly among cold-exposed young and old individuals, the rate of decline is faster in aged males (73). These data are supported by the findings of Sugarek (162), who reported that median times for a return to baseline oral temperature after ingestion of iced water was 30% slower in aged (>60 yr) humans as compared with individuals between the ages of 40 and 59 yr. The age-related decrease in cold-exposed rectal temperature may be dependent on gender. Rectal temperatures of younger (20-30 yr) and older women (51-72 yr) exposed to temperatures ranging for 10 to 20°C for 2 h do not differ significantly. Rectal temperatures of older men are significantly less than those of older and younger women and younger men (181). Although the current experimental evidence supports a suggestion of blunted CIT during aging, significant conflicting data preclude a definitive conclusion.

Research using animal models to investigate the effect of aging on CIT is limited to rodent models (6, 28, 40, 44, 51, 67, 79, 94, 107, 109-115, 149, 150, 152, 154, 155, 165, 167-170, 185). These investigations are, in general, consistent with observations in humans that CIT declines with advancing age (Table 2). With the exception of one investigation (44), reports before 1980 do not provide sufficient data on animal husbandry or life-span characteristics necessary to rule out the influence of disease. Thus caution should be used when interpreting the results of these investigations. Virtually every investigation performed after life span and housing conditions were standardized reports that aged mice and rats have significantly lower cold-exposed core temperatures than do young animals; that is, the preponderance of evidence in studies using laboratory rodents suggests that CIT declines with advancing age. In addition, the observation in humans that older females tend to maintain core temperature in the cold better than do males is supported by the animal work (51, 107, 109, 113).

Appropriate perception of ambient temperature is critical to initiating the physiological response to cold exposure. Age-related functional decrements in other sensory systems such as vision, hearing, and taste (116) imply that the ability to perceive cold may also be attenuated in the elderly (186). Elderly human subjects exposed to gradually declining temperatures below thermoneutrality and allowed to adjust the ambient temperature to a self-selected comfortable point tend to make the initial adjustment more slowly than do younger individuals (124). The slower adjustment occurs despite the fact that body temperature of the older group declines to a greater degree than that of the younger individuals. The elderly also appear to be less able to discriminate between a close range of ambient temperatures; that is, younger individuals can distinguish ambient temperatures of ~1°C, whereas some older subjects are unable to detect differences of up to 4°C (25). Aging rats given access to a variety of ambient temperatures prefer a temperature at which they experience a significantly larger decline in body temperature relative to younger animals (128).

	III. HEAT LOSS

Top Previous Next References

The thermoregulatory regions of the hypothalamus initiate the decline in heat loss from the body via vasoconstriction of the cutaneous vasculature upon exposure to a cold environment (65). Peripheral blood flow is shunted away from the skin toward deep body tissue, and conductive and convective heat loss from the skin's surface is reduced. The ability to reduce heat loss during cold exposure is dependent, to a large degree, on the reactive tone of the cutaneous vasculature but may also include the insulative properties of body fat. Age-related changes in the amount of body fat and the response of the cutaneous vasculature could result in increased heat loss and lead to diminished core temperature during cold exposure.

Heat conservation during periods of cold exposure is linked directly to efficient vasoconstriction. Observations in humans suggest that inefficient vasoconstriction response of the aged during cold exposure is related to increased heat loss. Vasoconstriction is estimated indirectly as the difference between core and skin temperature (181, 182) or changes in peripheral blood flow (68, 84, 140). Direct evaluation of age-related vasoconstriction during whole body cold exposure has not been reported. Rather, much of the current understanding regarding this relationship is suggestive. Enlargement in the vessel diameter and wall thickness is associated with an age-related increase of arterial wall stiffness (89). Increase in the arterial wall stiffness decreases the distensibility of the vessel wall that, in turn, results in resistance to constriction and the likelihood of increased heat loss. Although the mechanisms reflecting the increase in arterial wall stiffness are unknown, decreases in elastin and increases in the concentration of collagen cross-links of the peripheral vasculature are suggestive of age-related alterations in the vasoconstriction (118, 138). These investigations have been generally carried out using large vessels, e.g., aorta, isolated from rodents or in tissues from type II diabetic humans. It is not clear if the small vessels within the skin, those involved primarily in vasoconstriction response of the healthy elderly, show a similar distribution of collagen cross-links with aging.

Investigations evaluating cutaneous fingertip blood flow as an index of vascular constriction during cold exposure suggest an equal magnitude in the vasoconstriction response of young and old subjects (84, 140). However, the vasoconstrictor response occurs more slowly in the elderly and takes twice as long to attain the same level of vasoconstriction as observed in the younger subjects. Richardson et al. (140) found that although the cold-induced vasoconstriction occurs within the first minute in both young and old subjects, the response to the cold is not maintained in the elderly, and blood flow returns to precold levels.

Diminished response of smooth muscle to neural and hormonal stimuli may contribute to blunted or delayed vasoconstriction response of the elderly and can significantly affect the amount of heat loss during cold exposure (19, 34, 45, 99, 139, 172). The increase in peripheral blood flow in beagles (61) and humans (68) is generally less during -adrenergic stimulation in older versus younger subjects. This does not necessarily indicate that decreased vasoconstriction in the aging subject reflects altered adrenergic response. The difference in blood flow does not appear to reflect an alteration in -adrenergic sensitivity (60). Hogikyan and Supiano (68) report that decreased responsiveness of the smooth muscle -adrenergic receptor is compensated for by increased sympathetic nervous system activity. The net outcome is no significant alteration in -adrenergic-mediated vasoconstriction response during aging. Although further evaluation of possible age-related changes to arterial smooth muscle is warranted, it appears that diminished vasoconstriction responsiveness during cold exposure in the elderly reflects primarily increased arterial wall stiffness.

The amount of body fat may also contribute significantly to an effective defense against excessive heat loss in older individuals (30). The epidemiological and experimental evidence suggest that the amount and percentage of body fat increase with age (37). Some suggest that the greater amount of body fat in older versus younger subjects improves thermoinsulation and decreases heat loss during cold exposure (73, 181). This improvement in thermoinsulation is related to a reduction in the skin-to-ambient temperature thermal gradient; that is, older, fatter males exposed to temperatures between 10 and 20°C have significantly lower skin temperatures than do younger, leaner individuals (181). The reduction in the thermogradient did not, however, prevent a greater decrease in cold-exposed rectal temperatures of the older versus younger males. Although the older females had the highest percent of body fat and had no change in rectal temperature during cold exposure, Wagner and Horvath (181) concluded that the amount of body fat does not fully reflect differences in cold-exposed rectal temperatures. They observed that younger women had 12% more body fat than age-matched males, yet the rectal temperature of these women was significantly lower. These investigators suggest that age-related differences other than percent body fat must account for the difference between older males and females. The differing results between males and females may reflect the greater variability associated with the measurement of body fat observed previously in elderly versus younger adults (131). This suggestion is supported by the observation that significant correlation between body fat and rectal temperature exists even when weak and nonsignificant differences are found in the percent body fat among older and younger men (12). When the body fat variable is removed from regression analysis, older men have higher skin temperature, greater heat loss, and lower tissue insulation than do younger individuals; that is, these investigators suggest that greater body fat in older men represents only a correlational rather than causal relationship to reduced heat conservation. Moreover, older rats that have similar or greater body fat than younger animals consistently show lower cold-induced core temperatures (115). Although some evidence indicates that the amount of body fat may impact the conservation of heat during cold exposure, it appears that inadequate cutaneous vasoconstriction is more likely the cause of age-related increases in heat loss.

	IV. HEAT PRODUCTION

Top Previous Next References

The effect of aging on metabolic heat production during cold exposure is controversial. Investigations evaluating metabolic heat production in humans and in rodents have found a decrease (72, 112, 155, 170, 171, 184) or no difference (51, 112, 182) in cold-exposed oxygen consumption rates between young and old subjects. The apparent discrepancy among the results most likely reflects differences in study design that include degree and duration of cold exposure as well as the age and gender of the elderly subjects. Although experimental conditions can be controlled to a greater extent in studies using laboratory rodents as compared with humans, inconsistency of results concerning the effect of aging on cold-induced metabolic heat production persists. In the following section, the discussion focuses on the affect of aging on the individual components that contribute to overall increases in metabolic heat production during cold exposure.

Heat produced via shivering thermogenesis in skeletal muscle is important for maintaining homeothermy and is estimated to provide up to one-third of the total heat production during cold exposure (48). Shivering thermogenesis may also become more important during aging as the contribution of nonshivering heat production appears to decline (see sect. IIIB). Therefore, changes in skeletal muscle could potentially play a role in the decline of cold-exposed thermoregulation during aging. However, similar to the results of heat loss during cold exposure, few investigations have directly linked diminished skeletal muscle shivering thermogenesis to the age-related reduction in cold-induced thermoregulatory capacity. Rather, age-related declines in skeletal muscle mass and metabolic activity are cited as evidence for decreased heat production via shivering (82).

View larger version (40K): [in this window] [in a new window]   FIG. 2.   Total body mass-independent oxygen consumption (O2; mean ± SE) of 12- and 24-mo-old sedentary and exercise-trained male Fischer 344 rats during 6 h of cold exposure (6°C). * Significant difference from other 3 groups (P < 0.05). [From McDonald et al. (112).]

Brown adipose tissue is the principal anatomic location of nonshivering thermogenesis (63). The mechanisms that induce thermogenesis in this tissue have been identified and are discussed at length elsewhere (4, 64). Briefly, the release of norepinephine from the terminal endings of sympathetic neurons initiates the events leading to the onset of nonshivering thermogenesis. Norepinephrine binds primarily to -adrenergic receptors, including the novel adrenoceptor subtype ₃ (4), and the transduction of this signal is mediated through a second messenger pathway involving adenylyl cyclase. In turn, fatty acids are mobilized from intracellular stores of triacylglycerides and oxidized in the mitochondria. The oxidation of fatty acids is "uncoupled" from the electron transfer system via a novel uncoupling protein within the mitochondria so that heat, rather than ATP, is the metabolic end product (Fig. 3; Refs. 54, 173, 174). Although BAT constitutes only ~1% of the total body mass in the adult rodent, this tissue contributes as much as one-third to total cold-exposed oxygen consumption (48). Nonshivering thermogenesis in BAT can be an important component in the maintenance of homeothermy during cold exposure.

View larger version (18K): [in this window] [in a new window]   FIG. 3.   Schematic describing influence of norepinephrine (NE) on uncoupling of oxidative phosphorylation leading to brown fat nonshivering thermogenesis. 3, -adrenergic receptor; cAMP, adenosine 3',5'-cyclic monophosphate; HSL, hormone-sensitive lipase; UCP, uncoupling protein. [From Trayhurn (173).]

The role of BAT in human CIT is less clear. Several investigations have identified significant amounts of BAT in prenatal and newborn humans (1, 75). Because skeletal muscle function is not fully developed in the newborn, nonshivering thermogenesis is the principal mechanism for heat production in the postnatal infant. As the human matures, however, significant atrophy of BAT occurs. It remains to be elucidated if atrophy of BAT in humans reflects a normal biological aging process or the capacity of the human to adjust environmental factors as a means to protect against the cold (i.e., turn up the heat, put on a jacket). Nonetheless, autopsy investigations evaluating the content of BAT in adults have found depots of this tissue in similar anatomic locations as those observed in rodents (1, 74). To our knowledge, only one investigation has shown that blood flow, one index of metabolic capacity, increased to the perirenal BAT depot in response to norepinephrine, the physiological stimulus for nonshivering heat production (5). These data suggest that the response of human BAT to its principal agonist, norepinephrine, is intact. Moreover, mitchondrial uncoupling protein, the identifying characteristic of BAT, has been immunohistochemically identified in humans (87). Although the role of BAT nonshivering thermogenesis in humans has yet to be fully elucidated, it is conceivable that a reduction in this tissue's capacity to produce heat could contribute to the blunted cold-exposed thermoregulatory response of the elderly.

View larger version (31K): [in this window] [in a new window]   FIG. 4.   Interscapular brown adipose tissue (IBAT) mass (A), IBAT uncoupling protein (UCP) concentration (B), brown adipocyte cell size (C), and IBAT cell proliferation (D) in male Fischer 344 rats exposed to cold (10°C) or noncold (25°C) for 5 days. Non-cold-exposed means sharing a common lowercase letter do not differ significantly (P < 0.05). Cold-exposed means sharing a common uppercase letter do not differ significantly (P < 0.05). * Within an age group, cold-exposed value is significantly different from non-cold-exposed mean. [From Florez-Duquet et al. (47).]

View larger version (27K): [in this window] [in a new window]   FIG. 5.   Interscapular sympathetic neural activity and change in colonic temperature (TCO) in 10- and 30-mo-old C57BL/6J mice during acute cold exposure followed by a reheating period. [From Kawate et al. (79).]

View larger version (11K): [in this window] [in a new window]   FIG. 6.   -Adrenergic signal transduction pathway leading to an intracellular response. -Ar, -adrenergic receptor; Gs, stimulatory G protein; AC, adenylyl cyclase.

	V. COLD-INDUCED THERMOREGULATION AND THE RATE OF AGING

Top Previous Next References

This review has focused on possible mechanisms responsible for the age-related decline in CIT. Although these reports have provided substantial new information concerning the influence of age on thermoregulation, the physiological mechanisms have yet to be clearly identified. At least part of the problem in elucidating the causes of the age-related diminution in CIT reflects the inability to precisely define biological senescence. Most investigations reported here use chronological time to define aging. The use of chronological rather than biological time poses problems of interpretation because it assumes that age-related loss will occur to all animals at a specific time. However, the rate of aging is specific to the individual. For example, we generally use 24- to 26-mo-old F344 rats to evaluate the affect of age on CIT and have considered this group to be senescent. These ages are chosen because the mean life span of the F344 is ~24 mo. Mean core temperature is consistently lower in cold-exposed 26-mo-old male F344 rats compared with 6- and 12-mo-old animals. There is, however, significant heterogeneity in the response of the oldest rats to the cold exposure; that is, some rats do not display significant alterations in cold-exposed core temperature, whereas others develop severe hypothermia.

Close inspection of our data reveals a correlation between body weight loss and core temperature. The greater the body weight loss before the cold exposure, the greater the decline in cold-exposed core temperature. In previous investigations, rats showing marked body weight loss were eliminated from the study because we believed that they were "sick" and did not represent normal biological aging. However, when rats were examined pathologically to determine the cause of death, a consistent "disease" accounting for the well-characterized rapid loss in body weight near the end of life was not detected (123). These observations led us to hypothesize that a loss in body weight near the end of life may represent a transition from gradual aging to rapid senescence and may be a normal age-related phenomenon.

To test this hypothesis, we exposed rats to 6°C for 4 h every 14 days beginning at 24 mo of age and until the onset of rapid spontaneous weight loss (108). The age of onset of the rapid spontaneous weight loss ranged from 24.5 to 29 mo of age. All rats displaying spontaneous rapid weight loss had significant hypothermia during the acute cold exposure. The development of hypothermia in the spontaneous rapid weight loss group was not, in general, observed before their weight loss. Moreover, weight-stable rats of identical age maintained normal core temperature during cold exposure (Fig. 7). The rapid weight loss was associated with significant changes in body fat and protein, a 30% reduction in food intake, and lower levels of brown fat UCP that was not observed in the weight-stable animals.

View larger version (21K): [in this window] [in a new window]   FIG. 7.   Colonic temperature of weight-stable vs. weight-unstable old male Fischer 344 rats during resting conditions in a thermoneutral environment and during cold exposure (4 h at 6°C). * Significant difference from weight-stable group (P < 0.05). [Adapted from McDonald et al. (108).]

These data indicate that the diminished CIT of the aging F344 rat more closely reflects body weight loss rather than chronological age. Although it is premature to conclude that body weight loss is a biomarker of aging, these data point out that definitions of aging based on chronological markers may significantly affect data interpretation; that is, the use of chronological benchmarks as predictors of senescence are less than optimal and may cause significant heterogeneity within the data. Moreover, our data suggest that older rats of identical age and genetic backgrounds may be in different physiological states depending on their biological age.

With the use of CIT as a general model of aging, our data suggest that the transition from gradual aging to rapid senescence involves centrally mediated regulation. This area of regulation may include the hypothalamus. For example, we find that CIT and energy balance (body weight and food intake), two systems that involve hypothalamic regulation, appear to undergo rapid senescence at approximately the same time. Recently collected preliminary data support the suggestion that regions of the hypothalamus may be associated with "end-of-life" physiology. We evaluated endogenous circadian rhythm of body temperature (CRT), a function known to be regulated by the suprachiasmatic nuclei region of the hypothalamus (120), in rats from 23 mo of age until natural death. Measures of CRT, daily mean body temperature, rhythm period, and amplitude do not differ significantly within each rat during the time when body weight was stable. However, at the onset of rapid weight loss, daily mean temperature and rhythm amplitude decreases, and the strength of the CRT period is significantly less. These data are consistent with the suggestion that the hypothalamus may be an important regulator of the rate of aging.

	VI. SUMMARY AND CONCLUSIONS

Top Previous Next References