The question of whether caloric restriction (CR) is hormetic is addressed in terms of two common definitions of the term. In terms of the older definition, i.e., a growth-stimulatory effect when lower doses of a compound which resulted in a growth-inhibition at higher doses, CR is better characterized as a co-hormetic [i.e., a paradigm which at relatively "low doses", in combination with some stimulus, will evince increased growth (proliferation) and at higher "doses" will inhibit this increased proliferation] rather than a hormetic agent. Mechanisms such as cellular selection of cellular subpopulations, increases in receptor efficiency, and preservation of cellular proliferative potential can interact with agents and produce increased growth as long as the CR is not too severe.

In terms of a broader definition, i.e., non-monotonic dose-response behavior of a compound for any adverse response, CR appears to be hormetic, both as a result of body weight (BW) loss and other potential mechanisms. The impact of changes in BW, or frank CR, can be considered a component of every test for hormesis, and is thus capable for interaction with any other agent. The changes that BW loss (or CR) induce are so profound that any aspect of an agent's action ­ metabolism, pharmacokinetics, pharmacodynamics- can modulate the response of an organism to an agent.

Similarly, other effects of a chemical that induce BW loss, e.g., physical activity or temperature dysregulation, can also induce dose-response curves that appear hormetic. The interaction of the hormetic agents of BW loss and CR can influence agent test.

Controlling these factors may make it possible to dissect out the key components of a hormetic response. In addition, the effects of CR or BW loss appear to extrapolate well across species (e.g., Colman and Kemnitz, 1999). Thus there is some reason to believe that these hormetic factors may be important for humans, and may already be a factor for tests of potentially adverse agents already conducted in humans.

INTRODUCTION

The original definition of hormesis, which stimulated much interest in the area (Calabrese, 1994), was the presence of a growth-stimulatory effect when lower doses of a compound which resulted in a growth-inhibition at higher doses were given (Southam and Erhlich, 1943) (Definition 1). The term appears to have been extended to include non-monotonic dose-response behavior of a compound for any adverse response e.g., anti-carcinogenesis induced by low doses of a carcinogen (Furst, 1987; Boxenbaum, 1988) (Definition 2).

Caloric restriction (CR) has been discussed as an hormetic agent (Neafsey et al., 1989; Turturro and Hart, 1994; Turturro et al, 1998a) or its anti-aging action the consequence of an hormetic agent (hypercorticism) (Masoro, 1998). Caloric intake, and the body weight (BW) resulting from the consumption, is a factor in every long-term test. These parameters also have been found to significantly modulate both spontaneous disease (Turturro and Hart, 1992; Turturro et al., 1993-1998a,b) and induced toxicity (Allaben et al., 1990; Hart et al., 1996; Turturro et al., 1998c).

Thus the modulatory effects of caloric intake and BW potentially affect the dose-response relationship of any potential hormetic agent in a long-term experiment. Since BW loss, which also occurs with CR, often occurs with exposure to agent (in fact, the degree of BW loss was often used as part of the process of determination of the maximum tolerated dose in a carcinogenicity test [ Interagency Staff Group, 1986]) the relationship of CR, BW loss and hormesis is useful to elaborate in order to help the interpretation of the experimental tests for hormesis.

Definition 1

Background

By a literal reading of definition 1, caloric restriction (CR) by itself is not hormetic in animals. At low doses CR does not stimulate growth (i.e., BW). CR appears to inhibit BW in a monotonic, if nonlinear, fashion.

However, when growth is considered in a wider context, as growth of cells rather than the complete animal, a different picture emerges. Although CR directly results in inhibition of proliferation in different cell populations, at least at some ages in the lifespan (e.g., Lu et al., 1993; Wolf and Pendergrass, 1995), the consequences of this inhibition in different subsystems, and consequently CR, within the organism can be considered as hormetic under certain conditions. Some examples of physiological subsystems where this can occur are in the immune and reproductive systems, as well as in the liver.

Immune System

The immune system is one of an organism's subsystems most sensitive to nutritional manipulation in general, and CR in particular. Malnutrition, either protein-calorie (e.g., McCarter et al., 1998), mineral (Kukreja and Khan, 1998) or vitamin (Trakatellis et al., 1997) can quickly impair immune response. CR, undernutrition without malnutrition, in the presence of a stimulating agent, such as the plant lectins PHA and ConA, increased spleen cell reactivity 2- to 3- fold compared to ad libitum (AL) fed animals at the same ages (Weindruch et al., 1982). Similarly, the capacity of lymphocytes given ConA to produce IL-2 was stimulated by CR (Pahlavani et al., 1997). CR also stimulated Natural Killer Cell responses, in conjunction with a dose of an interferon inducer (Weindruch et al., 1983). In addition, older animals showed preserved immune function with CR in mice (Venkatraman et al., 1994) and rats (Fernandes et al., 1997), i.e. compared to controls they were able to respond better after a stimulus. These effects may be a result of CR altering the composition of cell subpopulations in vivo, selecting for populations with increased growth potential (Weindruch and Walford, 1988). Therefore, when an appropriate test stimulus, such as a plant lectin, is used, low "doses" of CR, i.e. "doses" below the high CR which cause protein-calorie malnutrition, will result in increased "growth", i.e., proliferation, when combined with the appropriate test stimulus. CR can thus be considered, similar to the concept of co-carcinogen (Williams and Weisburger, 1991), a co-hormetic for immune function in animals across all ages. A co-hormetic would then be an agent given at low "doses" which, when combined with another agent, would result in growth enhancement when compared to the use of either no "dose" or high "doses" of the co-hormetic.

Liver

Liver size fairly closely correlates with body weight in healthy animals (Adolph, 1949). Because of the loss of BW with CR, there is a loss in liver size, which appears to be accomplished by decreased proliferation and increased apoptosis (James and Muskhelishvili, 1994). The targets of the increased apoptosis appear to be hormone sensitive cellular subpopulations in the liver, which may contain the majority of the genetically damaged cells in the organ (Muskhelishvili et al., 1996).

Thus it would appear that CR has simply an inhibitory effect on growth in this organ. However, CR, if it is not so severe as to become protein-calorie malnutrition, in conjunction with hepatectomy, results in an elevated level of regrowth (Himeno et al., 1992). This appears to be a result of increasing receptor sensitivity to the autocrine factors involved in liver regeneration. The role of cell subpopulation selection, as demonstrated for mouse liver, has not been characterized in this model, however, it remains a strong possibility for the mechanism for what has been observed. Again, CR is a co-hormetic in liver, with the stimulatory process being partial hepatectomy.

Reproductive System

In most commonly used experimental rodent models, reproductive potential in females is limited to early periods in the lifespan. For instance, female breeders are often retired before 11 months of age, developing various forms of anestrous after that time. In mice, CR has been found to inhibit early reproduction and cycles (Nelson et al., 1985). Upon re-feeding, however, animals as old as 33 months can cycle. When compared to similar aged AL animals, the AL animals can be considered not capable of being stimulated by endogenous gonadotropic hormones, while the CR animals can respond to both exogenous and endogenous hormones. This is probably accomplished by preserving cells that would normally have died through the cycles of estrous inhibited by CR. Again, CR is enriching a particular subpopulation of cells (oocytes) in the ovary, allowing them to respond to either endogenous or exogenous hormones. CR appears to be a co-hormetic in this subsystem, allowing the oocytes to maintain their sensitivity to endogenous stimulatory factors.

Summary - Definition 1

The effect of CR on preserving the functional and replicative potential of cells, as it does in the female reproductive system in mice, may be a general occurrence in the proliferative compartments of a number of organ subsystems.

Wolf and Pendergrass (1995) found that CR inhibits cellular proliferation up to ten months of age in cells derived from a number of different organs. This inhibition will result in the ability of cells to proliferate later in life, upon refeeding, in older animals, as supported by the observation of increased replicative potential in older cells. This potential will be observed when a stimulus is applied which results in increased proliferation.

The effect of CR on selecting cell subpopulations is not as well characterized in other organ systems, except for the preservation of cells that have the capacity to reproduce. Cells dependent upon those hormonal stimuli that are inhibited by CR to proliferate (such as Leydig cells in the testis, insulin-producing cells in the pancreas, various hormonally sensitive cells in the pituitary [such as somatotrophs]) will have inhibited replication in the CR animal, and the composition of the organs will reflect this. The functional significance of this has been demonstrated by the finding of different Cytochrome P-450 isoforms in testis of F-344 rat with CR, with direct consequences for the formation of tumors in that system (Seng et al., 1996). The selection for cells which do not require the hormonal stimuli inhibited by CR will again be evident when a stimulus is given that interacts with those cells.

The alteration of the response of receptors by CR is another interesting mechanism for the co-hormetic effects of CR. Changes in insulin sensitivity (Ivy et al, 1991) and the receptors involved in intermediary metabolism are examples of this. In addition the changes in the antioxidant defense systems (Sohal and Weindruch, 1996) may also reflect receptor changes in these systems. These effects become evident only when the receptor ligand is used, whether it is insulin or oxidative damage.

It appears that CR can act as a co-hormetic by either altering cell populations or altering cell receptors. Additionally, given the many effects CR has on metabolism and compound disposition, CR may also alter metabolic and excretory parameters for growth-stimulatory compounds. Thus, many mechanisms exist for the induction of co-hormetic interactions by CR.

Definition 2

Background

The second definition of hormesis is much broader than the first. It seems to include any non-monotonic dose-response relationship for any adverse effect. It is usually thought of as a "U-shaped" dose-response curve, i.e. the effect at zero dose is some value, which goes down with increasing dose to a minima, and then rises. Figure 1 demonstrates such a relationship of mortality to dose at different times-on-test (ages) in a two-year chronic test of a compound (benzyl acetate) (NTP, 1993).

Figure 1. Mortality in B6C3F1 Male Mice at Two Different Times-On-Test As a Function of the Dose of Benzyl Acetate. Mortality ( in percent) as a function of dose of Benzyl Acetate (in parts-per-million). Cohort is originally 50 male B6C3F1 mice at ages of 20 months (18 months-on-test plus test onset at two months of age) and 26 months (24 months-on-test plus test onset at two months of age) (terminal sacrifice). Note "U-shaped" dose-response curve, which appears to have a minima at the middle dose at 18 months-on test, and a minima at the low dose at terminal sacrifice. Mortality data from NTP (1993).

Although there are only four data points per time-on-test, it appears that the compound is hormetic, and that the effect is increasingly evident as the animals get older. Survival in the control group (zero dose) keeps deteriorating with age while mortality is inhibited at the medium doses of compound. In addition, the dose that exhibits the hormetic effect appears to get smaller as the animal ages.

CR has a hormetic relationship with survival (Turturro and Hart, 1994). A zero "dose" results in no increased survival or lessened tumor incidence when compared to AL feeding, while high "doses" of CR result in protein-calorie malnutrition. This is difficult to measure directly, since there are few experiments that expose animals to long-term starvation. And long-term experiments in which survival was poor in control animals would probably be terminated. However, using BW2 (body weight at 2 months-on-test [four months of age since the tests are usually started at two months of age]) as a surrogate, survival can be related to BW early in the lifespan for a number of chronic studies. This is illustrated for survival in Figure 2.

Figure 2. Mortality in B6C3F1 Male Mice As a Function of Body Weight at 2 Months-on-test. Mortality ( in percent) as a function of body weight at 2 months-on-test (BW2) (4 months of age [2 months-on-test plus test onset at two months of age]). Data on descending arm from three month sub-chronic bioassays from Technical Reports (similar to NTP, 1993) TR 299, 333, 345, 355, 387, 412, 442, 452. Data on the ascending arm are from chronic bioassays for feed studies (male mice) and are reported in Turturro et al. (1996).

This figure relates the percentage mortality in a cohort of approximately 50 animals in a 2-year bioassay (Turturro et al., 1996) and mortality in the short-term tests (3-month) used to generate the information to dose animals in the 2-year tests. Three-month studies were used because of the paucity of data from long-term tests with poor survival (see above). Also, wider dose ranges of chemical were used than in long-term tests and sometimes these exposures resulted in very low BW. Although these data could be considered complicated by the presence of the agents that induced the low BW, the animal responses (little evident pathology, proportionate organ weight changes, etc.) suggested that the induced low BW rather than some chemical-specific action was the most likely candidate for the agent causing death in these cases. If results from 2-year tests for this endpoint were available in this range, it is likely that the 2-year tests would be more sensitive to low BW than the three-month tests, so the slope of this segment of the curve would probably be steeper.

Although any hormetic effect of CR for tumors is not as clear as on mortality, an effect can be seen for some neoplasms. One example, liver tumors in male mice, is illustrated in Figure 3 (Seilkop, 1995).

Figure 3. Probability of Liver Tumor As a Function of Body Weight at 1 month-on-test (BW1). Liver tumor (adenoma plus carcinoma) probability in BW cohorts (1 gram per point, at least 10 animals per point) of B6C3F1 male mice as a function of BW1 (3 months of age [1 month-on-test plus test onset at two months of age]).

The very low early BW may indicate problems associated with cage fighting as dominance is established. The smaller males may not get adequate access to food. Lack of proper food consumption could lead to malnutrition, and certain deficiencies, such as folate deficiencies have been associated with increased tumors (Poirier, 1994).

In addition to its effects on BW, CR may have some other effects, e.g., related to hypercorticism. These changes would provide new possibilities for interaction with agents to result in hormesis

It is fairly clear then that CR is hormetic in itself using Definition 2 for morality, and perhaps other endpoints.

Interactions

Since caloric intake (or its consequent BW) is hormetic in effect, the interaction of CR with another agent that induced a reduction in BW at critical ages could look like a hormetic agent if the positive effect on a parameter such as lifespan from the BW reduction outweighed the adverse effects of low doses of the compound. This may be occurring in the case of benzyl acetate. Figure 4 is the same as Figure 1 with the addition of the BW2 of these animals.

Figure 4. Mortality and BW2 in B6C3F1 Male Mice at Two Different Times-On-Test As a Function of the Dose of Benzyl Acetate. Repeat of Figure 1 with additional characterization of BW2 (4 months of age[2 months-on-test plus test onset at two months of age]). Note the decline of BW2 with dose. Data from NTP (1993).ages are here to work around the problem.

The lowered BW2 (in a dose-response manner) are consistent with the improved survival seen at the low doses (Turturro and Hart, 1994, Turturro et al., 1996). It appears that at the higher dose (18 months) or higher doses (24 months), the toxicity of the compound is able to overcome any positive effect on mortality of lowering BW2.

As discussed elsewhere (Turturro and Hart, 1994), this confounding of toxicity effects with the consequences of reduced BW can be subtle. For instance, an agent can induce a decrease of BW2 and a BW increase at a later age. The decrease in BW2 is consistent with an improvement in overall survival, while a BW increase later in the lifespan can indicate higher incidences of certain tumors (Turturro et al., 1996, 1998a). Also, the magnitude of the increase in BW late in the lifespan may be greater than that of the decrease early. This could lead to an incomplete characterization of the effect of exposure to the agent as "increasing BW and increasing survival" while the treatment is actually only increasing BW at certain times in the animal's lifetime while increasing lifespan.

Thus, a long-term test can be considered to be not the test of a single compound, but of two agents, induced BW effects and a chemical. These two agents can have diametrically opposed effects on an adverse endpoint, such as mortality, as occurs with benzyl acetate. The two different mechanisms, a BW loss than decrease mortality and a chemical induced toxicity that increases it, can combine to produce a non-monotonic dose response curve.

Additionally, as an hormetic agent, CR is potentially capable of synergism, additivity and antagonism with other agents on almost any adverse endpoint. This is especially true because of the many aspects of the organism affected by CR, from water intake (Duffy et al., 1995), excretion and endogenous nutrient levels (e.g., ascorbate levels) (Taylor et al., 1995) to DNA repair (Lipman et al., 1989), oncogene expression (Hass et al., 1993) and cellular proliferation (Wolf and Pendergrass, 1995). Each effect has the potential of changing agent disposition, metabolism and/or pharmacodynamics to make the agent less bioactive (Hart et al., 1995; Turturro et al., in press). These modulatory effects can be overwhelmed when the agent achieves a certain concentration. When the major mechanism of toxicity by an agent is similar to some disease process occurring in the animal, then the consequence may be a hormetic effect. For example, a compound's major toxic effect may result from the stimulation of Thyroid Stimulating Hormone (TSH) through disruption of a part of the homeostatic control of TSH level (e.g., such as F.D.& C. Red Dye Number 3 [Hart et al., 1986]). Induced BW loss appears to decrease the levels of TSH required by the animal (Han et al., 1998), e.g., CR reduces body temperature (Duffy et al., 1995). The lowering of TSH requirements as a result of induced BW loss may be greater than the increases in TSH induced by the chemical up to a certain dose of chemical. The dose-response curve would then indicate that the chemical is having an hormetic action.

Consequences of BW Changes

Many of the effects of CR related to hormesis appear to be mediated by BW and BW growth patterns, as noted above. For example, the decrease in proliferation and increase in apoptosis seen with CR is correlated to individual animal BW in male mouse liver (Figure 5).

Figure 5. Apoptotic Bodies (AB) and PCNA+ Measurement in Livers of Individual 12-Month Old Male B6C3F1 Mice. Percentage incidence of apoptotic bodies (AB) (a measure of apoptosis) and PCNA+ (a measure of proliferation) in livers from individual 12-month old male B6C3F1 mice. BW is body weight at this age in grams. (r) is the linear (in log) correlation coefficient for both relationships. Techniques (and animals) are same as reported in James and Muskhelishvili, 1994.

The consequences of this relationship are that other physiological chemically-induced changes, such as increased activity, which modulate BW may also induce hormesis. Models to understand the effects of increased physical activity include exercise and cold stress.

Whether exercise can extend lifespan (i.e., either decrease mortality acceleration with age or change the intercepts of the survival curve) is a contentious issue. Older studies have shown an increase in mean lifespan (Edington et al. 1972; Drori and Folman, 1976; Sperling et al., 1978) as well as a decrease (Del Maestro, 1980) with exercise. There has also been shown to be an increase in maximum lifespan in two different rat strains (Retzlaff et al., 1966; Goodrick, 1980). Holloszy et al. (1985) has found little effect on maximum survival in voluntary exercising in another rat strain.

The issue is also seemingly clouded for humans, in that individuals who participate in exercise (e.g., Cambridge oarsmen and other athletes) outlive the general populace, but not their classmates (Rook, 1954; Montoye et al., 1956). Danish athletic champions outlive their contemporaries before age 50, and die faster after 50 (Schnohr, 1971).

One way to make sense of this information is take the lesson from studies using CR that the protocols used may be as important as the use of the technique itself. The studies which have shown some positive effect of exercise on survival in rats have tended to start the exercise early in life (1 ­ 4 months of age), while exercise started later seems to have little effect or even can be adverse (Goodrick, 1974). This indicates that timing and level of exercise, similar to what occurs with the use of CR, may be important. Previous work has shown that exercise increases the cardiac weight/body weight ratio by 50% (Oscai and Holloszy, 1970). In addition, experiments using cold stress, which had the advantage of careful pathological examination, showed a tenfold increase in the incidence of severe myocardial fibrosis and indications of an increase in periarteritis (Holloszy and Smith, 1986). If these are the effects that are related to any exercise-increased mortality, they will be very difficult to assess in a cause-of-death analysis since they are present in every aging animal in some degree. Cause-of-death analyses has been quite problematic even for two-year rodent chronic tests (Squire, 1988; Haseman et al, 1994).and become even more confounded in older rats since they have so many diseases at once that ascribing death to a single disease is often subjective.

However, it appears that exercise induces an hormetic agent which can interact with other exercise-induced physiological changes. The effect on survival of exercise and BW can be:

• 1) antagonistic, completely, as was highlighted by Holloszy and Schechtman (1991) who demonstrated that the combination of dietary restriction DR and exercise resulted in a significant increase in mortality in the first half of the lifespan (A later attempt by Holloszy to replicate this result [Holloszy, 1997] failed); or partially, by Holloszy et al. (1985) in which an exercise-induced 33% lower BW resulted in much less effect than a pair-weighed control; or

• 2) additive, as shown by Holloszy (1997), in which survival with DR was not effected by exercise, and an approximate 25% exercise- induced lowered BW significantly improved average survival, and was proportionately less effective than a DR-induced 45% lowered BW.

A similar phenomenon occurs when using CR. In Sprague-Dawley rats, the same 35% CR can either not significantly effect mammary tumor growth Keenan and Soper, 1995), or arrest it completely (Kritschevsky and Klurfield, 1987). In this circumstance, the keys to understanding the effects are the baseline control BW. When the AL control is a one kg animal, a 35% DR, which brings an animal to approximately a 650 g BW, CR has little effect on mammary tumor incidence when compared to controls. When the CR animal is 250 g, both spontaneous and induced mammary tumorigenesis can be inhibited completely (Kritschevsky and Klurfield, 1987).

For the exercised animals, when the CR pair-fed animal is approximately 400 g. (Holloszy et al., 1985), exercise is antagonistic to the BW effect, but when the CR animal is 325 g (Holloszy, 1997) not even exercise can inhibit the BW effect. The relationships of body size to tendency to neoplasia or survival are not linear and a small animal is much more refractory to tumorigenesis than a large one, even if the large one is CR. Thus it appears that exercise can induce some of the effects of CR on survival, but can also have adverse effects.

On the other hand exercise slows the age associated loss of lean muscle mass per gram body weight (Garthwaite et al., 1986) and help make bone denser (Pollock et al., 1997). The effect on bone is different than the reduction in density that occurs with CR-induced BW loss (Sanderson et al., 1997).

Thus, when interacting with another agent, the BW changes induced by exercise are consistent with extended lifespan, some of the exercise effects appear to be deleterious to survival, while other exercise effects provide benefits that BW loss does not confer. This is similar to potential effects of physiological changes induced by CR that may not completely be the result of BW loss. These provide additional avenues for interaction with chemical actions. Thus, the multi-component paradigm exercise can have some of the same overall consequences as BW loss for hormesis, i.e., decrease the occurrence of adverse consequences until overtaken by the toxic effects of high doses of a compound. This is especially important since increases in activity often occur when some compounds, such as amphetamine, are given and these increases in activity may have many of the same consequences as exercise.

Conclusion

It appears that CR, in terms of the older definition of hormesis (Definition 1) is better characterized as a co-hormetic (i.e., a paradigm which at relatively "low doses", in combination with some stimulus, will evince increased growth (proliferation) and at higher "doses" will inhibit this increased proliferation) rather than a hormetic agent. Mechanisms such as cellular selection of cellular subpopulations, increases in receptor efficiency, and preservation of cellular proliferative potential can interact with agents and produce increased growth as long as the CR is not too severe. What is the optimal level of CR for this co-hormesis may be dependent on the organ system and cell population studied. CR can mitigate specific cancers and specific functional aging losses in a co-hormetic fashion. An approach that investigates this question in detail is more likely to assist the application of CR as a co-hormetic in practical situations that an approach that discusses the complex and multi-focal phenomena called aging or cancer.

In terms of the broader definition (Definition 2), CR appears to be hormetic, both as a result of BW loss and other potential mechanisms. The impact of changes in BW, or frank CR, can be considered a component of every test for hormesis, and is thus capable for interaction with any other agent. The changes that BW loss (or CR) induce are so profound that any aspect of an agent's action ­ metabolism, pharmacokinetics, pharmacodynamics- can modulate the response of an organism to an agent.

Similarly, other effects of a chemical that induce BW loss, e.g., physical activity or temperature dysregulation, can also induce dose-response curves that appear hormetic.

The interaction of the hormetic agents of BW loss and CR can influence agent test. Controlling these factors may make it possible to dissect out the key components of a hormetic response. In addition, these factors provide another tool to manipulate biological systems in order to understand the fundamental mechanisms that underlie health and disease. The effects of CR or BW loss appear to extrapolate well across genotype (Sheldon et al., 1996) and species (e.g., Colman and Kemnitz, 1999). Thus there is some reason to believe that these hormetic factors may be important for humans, and may already be a factor for tests of potentially adverse agents already conducted in humans.

REFERENCES

Adolph, E. Quantitative relations in the physiological constitutions of mammals. Science 109:579-585, 1949.

Boxenbaum, H. , Neafsey, P., Founier, H. Hormesis, gompertz functions, and risk assessment. Drug Metabol. Reviews 19:195-229, 1988.

Calabrese, E. Primer on BELLE. In: Biological Effects of Low Level Exposures: Dose-Response Relationships (Edward Calabrese, Ed.) Lewis Publishers, Chelsea, MI 48118, pp. 27 - 42, 1994.

Colman R, JW Kemnitz. Aging experiments using nonhuman primates. In: Methods in Aging Research (B.P. Yu, ed.) CRC Press, Boca Raton, FL,pp. 249-267, 1999.

Del Maestro, R. An approach to free radicals in medicine and biology. Acta Physiolog. Scand. 492:153-168, 1980.

Drori, D., Folman, Y. Environmental effects on exercise in the male rat: exercise, mating, castration and restricted feeding. Exp. Gerontol. 11:25-32, 1976.

Duffy, P., Feuers, R., Pipkin, J. Berg, T.F., Divine, B., Leakey, J., Turturro, A., Hart, R. The effect of caloric modulation and aging on the physiological response of rodents to drug toxicity. In Dietary Restriction: Implications for the Design and Interpretation of Toxicity and Carcinogenicity Studies (R. Hart, D. Neuman, and R. Robertson, Eds.) ILSI Press, Wash., DC. ppg. 127-140, 1995.

Edington, D., Cosmas, A., McCafferty, W. Exercise and longevity: evidence for a threshold age. J. Gerontol. 27:341-3, 1972.

Fernandes, G., Venkatraman, J., Turturro, A., Attwood, V., and Hart, R. Effect of food restriction on lifespan and immune functions in long-lived Fischer-344 X Brown Norway F1 rats. J. Clin. Immun. 17:85-95, 1997.

Furst, A. Hormetic effects in pharmacology: Pharmacological interventions as prototypes for hormesis. Health Phys. 52:527-530, 1987.

Garthwaite S, Cheng H, Bryan J, Craig B, Holloszy J. Ageing, exercise and food restriction: effects on body composition. Mech Age Develop. 36:187-96, 1986.

Himeno, Y., Engleman, R., Good, R. Influence of calorie restriction on oncogene expression and DNA synthesis during liver regeneration. Proc. Nat. Acad. Sci. 89:5497-5501, 1992.

Goodrick, C. Effects of long-term voluntary wheel exercise on male and female Wistar rats 1. Longevity, body weight and metabolic rate. Gerontology 26:22-33, 1980.

Han, E., Lu, D., Nelson, J. Food restriction differentially effects mRNA's encoding the major anterior pituitary tropic hormones. J. Gerontol. 53:B322-29, 1998.

Hart, R.W. Freni, S.C. Gaylor, D.W. Gillette, J.R. Lowry, L.K. Ward, J.M. Weisburger, E.K. Lepore, P. and Turturro, A. Final report of the color additive scientific review panel. Risk Analysis 6:117-154, 1986.

Hart, R., Keenan, K., Turturro, A., Abdo, K., Leakey, J., Lyn-Cook, B. Symposium Overview. Caloric restriction and toxicity. Fundamental and Appl. Toxicol. 25:184-195, 1995.

Haseman, J. Eustis, S., and Ward, J. Contributing causes of death and the use of this information in the statistical evaluation of tumor data. In : Pathology of the Aging Rat. U. Mohr, D. Dungworth, C. Capen, eds. ILSI Press, Wash. D.C., ppg. 629-638, 1994.

Hass, B., Hart, R., Lu, M.H., Lyn-Cook, B. Effects of caloric restriction in animals on cellular function, oncogene expression, and DNA methylation in vitro. Mutation Research 295:281-289, 1993.

Holloszy, J. Mortality rate and longevity of food-restricted exercising male rats: a reevaluation. J. Appl. Physiol. 82:399-403, 1997.

Holloszy, J., Smith, E. Longevity of cold-exposed rats: a reevaluation of the "rate-of-living theory". J. Appl. Physiol. 61:1656-60, 1986.

Holloszy, J.. Schechtman, K. Interactions between exercise and food restriction: effects on longevity of male rats. J. Appl. Physiol. 70:1529-35, 1991.

Holloszy, J., Smith, E., Vining, M., Adams, S. Effect of voluntary exercise on longevity of rats. J. Appl. Physiol. 59:826-31, 1985.

Interagency Staff Group Chemical carcinogens: A review of the science and associated principles. Environ Health Perspect 67:201-282, 1986.

Ivy, J., Young, J., Craig, B., Kohrt, W., Holloszy, J. Ageing, exercise and food restriction: effects on skeletal muscle glucose uptake. Mech. Age. Develop. 61:123-33, 1991.

James, J., Muskhelishvili, L. Rates of apoptosis and proliferation vary with caloric intake and may influence incidence of spontaneous hepatoma in C57Bl/6XC3Hf1 mice. Cancer Res. 54:5508-5510, 1994.

Keenan, K., and Soper, K. The effects of overfeeding and moderate dietary restriction on Sprague-Dawley rat survival, spontaneous carcinogenesis, chronic disease and the toxicologic response to pharmaceuticals. In: Dietary Restriction: Implications for the Design and Interpretation of Toxicity and Carcinogenicity Studies, (Hart, R., Neumann, D., and Robertson, R., Eds.) ILSI Press, Washington, D.C. pp. 99-126, 1995.

Kritchevsky, D., Klurfield, D. 1987. Caloric effects in experimental mammary tumorigenesis. Amer. J. Cancer Nutr. 45:236-242, 1987.

Kukreja, R., Khan, A. Effect of selenium deficiency and its supplementation on DTH response, antibody forming cells and antibody titre. Ind. J. Exp. Biol. 36:203-5, 1998.

Lipman, J., Turturro, A., Hart, R. The influence of dietary restriction on DNA repair in rodents: A preliminary study. Mech. Age. Devel. 48:135-143, 1989.

Lu, M., Hinson, W., Turturro, A., Sheldon, W., Hart, R. Cell proliferation by cell cycle analysis in young and old dietary restricted mice. Mech. Age. Develop. 68:151-162, 1993.

Masoro, E. Hormesis and the anti-aging action of caloric restriction. Exp.Gerontol. 33:61-66, 1998.

McCarter, M., Naama, H., Shou, J., Kwi, L., Evoy, D., Calvano, S., Daly, J. Altered macrophage intracellular signaling induced by protein-calorie malnutrition. Cellular Immunol. 183:131-6, 1998.

Montoye, H., Van Huss, W., Olson, H., Hudec, A., Mahoney, E. 1956. Study of the longevity and morbidity of college athletes. J. Amer. Med. Assn. 162:1132-4.

Muskhelishvili, L., Turturro, A., Hart, R., James, S. Jill. p-class glutathione-s-transferase positive hepatocytes in aging B6C3F1 mice undergo apoptosis induced by dietary restriction. Am J. Path. 149:1585-1591, 1996 .

Neafsey, P., Boxenbaum, B., Ciraulo, D., Fournier, D. A Gompertz age-specific mortality rate model of aging: Modification by dietary restriction in rats, Drug Metabol. Rev., 21, 351, 1989.

Nelson, J., Gosden, R., Felicio, L. Effect of dietary restriction on estrous cyclicity and follicular reserves in aging C57Bl6/J mice. Biol. Reprod. 32:515-22 , 1985.

NTP (National Toxicology Program) Toxicology and Carcinogenesis Studies of Benzyl Acetate, NTP Technical Report (TR) 431, National Institute of Environmental Health Sciences, Research Triangle Park, N.C., 1993.

Oscai, L., and Holloszy, J. Weight reduction in obese rats by exercise or food restriction: effect on the heart. Amer. J. Physiol. 219:327-30, 1970.

Pahlavani, M., Harris, M., Richardson, A. The increase in the induction of IL-2 expression is correlated to changes in the transcription factor NFAT. Cellular Immunol. 180:10-19, 1997.

Poirier, L. Methyl group deficiency in hepatocarcinogenesis. Drug Metabolism Reviews 26:185-99, 1994.

Pollock, M., Mengelkoch, M., Graves, J, Lowenthal, D., Limacher, M., Forster, C., Wilmore, J. Twenty-year follow-up of aerobic power and body composition of older track athletes. J. Appl. Physiol. 82:1508-16, 1997.

Retzlaff, E., Fontaine, J. Furuta, W. Effect of daily exercise on lifespan of albino rats. Geriatrics 21:171-7, 1966.

Rook, A. An investigation into the longevity of Cambridge sportsmen. Br. Med. J. 1: 773-7, 1954.

Sanderson, J.P., Binkley, N., Roecker, E.B., Champ, J.E. Pugh, T.D., Aspnes, L., Weindruch, R. Influence of fat intake and caloric restriction on bone in aging male rats. J. Gerontol. Series A: 52:B20-25, 1997.

Schnohr, P. Longevity and the causes of death in male athletic champions. Lancet II:1364-66, 1971.

Seilkop, S. The effect of body weight on tumor incidence and carcinogenicity testing in B6C3F1 mice and F-344 rats. Fund. Appl. Toxicol. 24: 247-259, 1995.

Seng, J., Gandy, J., Turturro, A., Lipman, R., Parkinson, A., Hart, R.W., and Leakey, J.E. Effects of dietary restriction on expression of testicular cytochrome P450 isoforms associated with the metabolic activation of carcinogens. Archives Biochem. Biophys. 335:42-52, 1996.

Sheldon, W., Bucci, T., Blackwell, B., and Turturro, A. Effect of ad libitum feeding and 40% feed restriction on body weight, longevity, and neoplasms in B6C3F1, C57Bl6, and B6D2F1 Mice. In: (U. Mohr, D. Dungworth, C. Capen, W. Carlton, J. Sundberg, and J. Ward, eds.) Pathobiology of the Aging Mouse, Vol. 1, ILSI Press, Wash. DC. ppg. 21-26, 1996.

Sohal, R., Weindruch, R. Oxidative, stress, caloric restriction and aging. Science 273:59-63, 1996.

Southam, C., Ehlich, J. Effect of extract of western red cedar hardwood on certain wood decaying fungi. Phytopathology 33:517-524, 1943.

Sperling, G., Loosli, J., Lupien, P., McCay, C. Effect of sulfamerazine and exercise on life span of rats and hamsters. Gerontology 24:220-224, 1978.

Squire, R. Criteria for classifying neoplasms and use of data on non-neoplastic lesions. In Carcinogenicity: the Design, Analysis and Interpretation of Long-Term Animal Studies. H. Grice, J. Cominera eds. Springer-Verlag, N.Y. pg. 107- 115, 1988.

Taylor, A., Lipman, R., Jahngen-Hodge, J., Palmer, V., Smith, D., Padhye, N., Dallal, G.E., Cyr, D.E., Laxman, E., Shepard, D., Morrow, F., Salomon, R., Perrone, G., Asmundsson,G., Meydani, M., Blumberg, J., Mune, M., Harrison, D.E., Archer, J.R., Shigenaga, M. Dietary calorie restriction in the Emory mouse: Effects on lifespan, eye lens cataract prevalence and progression, levels of ascorbate, glutathione, glucose, and glycohemoglobin, tail collagen breaktime, DNA and RNA oxidation, skin integrity, fecundity and cancer. Mech. Age. Devel. 79, 33-57, 1995.

Trakatellis, A., Dimitriadou, A., Trakatelli, M. Pyridoxine deficiency: new approaches in immunosuppression and chemotherapy. Postgrad. Med. J. 73:617-622, 1997.

Turturro A., Hart, R. Dietary alteration in the rate of cancer and aging. Experimental Gerontology 27:583-592, 1992.

Turturro, A., Hart, R. Modulation of toxicity by diet: Implications for response at low-level exposures. In: Biological Effects of Low Level Exposures: Dose-Response Relationships (Edward Calabrese, Ed.) Lewis Publishers, Chelsea, MI 48118, Chapter 9, pp. 143-152, 1994.

Turturro, A., P. Duffy and Ronald Hart. Modulation of toxicity by diet and dietary macronutrient restriction. Mutation Res. 295: 151-164, 1993.

Turturro A, Blank K, Murasko D, Hart R. Mechanisms of caloric restriction effecting aging and disease. Annals N Y Acad Sci. 719:159-170, 1994.

Turturro A, Duffy P, Hart R. The effect of caloric modulation on toxicity studies. In: Hart R, Neuman D, Robertson R, eds. Dietary Restriction: Implications for the Design and Interpretation of Toxicity and Carcinogenicity Studies. Wash., D.C.:ILSI Press; pp. 79-98, 1995.

Turturro, A., Duffy, P., Hart, R., and Allaben, W. Rationale for the use of dietary control in toxicity studies - B6C3F1 mouse. Toxicologic Pathology 24:769-775, 1996

Turturro, A., Leakey, J., Allaben, W., and Hart, R. Letter to the Editor, Response to Michael Festing's "Fat Rats". Nature 389:326, 1997.

Turturro, A., Hass, B. Hart, R. Hormesis ­ Implications for risk assessment caloric intake (Body weight ) as an exemplar. Human and Ecological Risk Assessment 17:454 - 499, 1998a.

Turturro A, Hass B, Hart R, Allaben W. Body weight impact on spontaneous and agent-induced diseases in chronic bioassays. Inter J Toxicol. 17:79-100, 1998b.

Turturro, A., Leakey, J., Allaben, W. And Hart, R. Modulation of the carcinogenic response by caloric restriction. In:(C. Ioannides, ed.) Nutrition and Chemical Toxicity. J. Wiley and Sons, Sussex, UK. ppg. 201-18, 1998c.

Turturro, A., Duffy, P. and Hart, R. in press. Antioxidation and evolution: Dietary restriction and alterations in molecular processes. T. Basu, N. Temple., and N. Garg (eds.) Antioxidants in Human Health and Disease. CAB International, Oxford, Eng.

Venkatraman, J., Attwood, V., Turturro, A., Hart, R., and Fernandes, G. Maintenance of virgin T-Cells and immune functions by food restriction during aging in long-lived B6D2F1 female mice. Aging:Immun. & Immunopath. 5:13-26, 1994.

Weindruch, R., and Walford, R. The Retardation of Aging and Disease By Caloric Restriction. C.C. Thomas, Springfield, IL, 1988.

Weindruch, R., Gottesman, S., Walford, R. Modification of age-related immune decline in mice dietary restricted from or after midadulthood. Proc. Nat. Acad. Sci. 79:898-902, 1982.

Weindruch, R., Devens, B., Raff, H., Walford, R. Influence of dietary restriction and aging on natural killer cell activity in mice. J. Immunol. 130:993-6, 1983.

Williams, G. Weisburger, J. Chemical carcinogenesis. In: Casseret's and Doull's Toxicology The Basic Science of Poisons (Amdur, M., Doull, J., Klaasen, C., eds.) New York: McGraw-Hill ppg.127-200, 1991.

Wolf N, Pendergrass W. In vivo cell replication rates and in vitro replication capacity are modulated by caloric intake. In: Dietary Restriction and Its Effect on the Design and Conduct of Carcinogenicity Studies, (Hart R, Neumann D, Robertson R, eds.) Wash, DC:ILSI Press ppg. 299-310, 1995.