Up to this point we have considered ways in which an organism can avoid the deleterious effects of the pathogen, either reflexively or by learning about pathogen-associated stimuli. A remaining question is what happens to behaviour as a result of infection and does this behaviour assist the organism to rid itself of the pathogen and return to a healthy state. As indicated above, there is a well-characterized profile of behavioural change observed in sick animals, called sickness behaviour (Table I). This profile of behavioural change includes decreased general activity (especially locomotor activity), decreased food intake [32-34], increased slow-wave sleep [35,36], altered sexual activity , decreased environmental  and social exploration , impaired cognitive function [40,41], and decreased interest in pleasurable stimuli (so-called "anhedonia") , As mentioned previously, Hart  argued that the typical profile of behavioural change experienced by organisms infected with a pathogen is a strategy used by the organism to aid return to health. Thus an important question is whether this sickness behaviour profile is a part of behavioural defence.
In his original analysis, Hart described the behavioural changes induced by pathogens (e.g., reduction in general motor activity, grooming and food intake) as being adaptive by assisting the organism to maintain its "emergency thermoregulatory mode" [2, p. 124]. For example, he argued that decreased mobility facilitates fevers because of the lack of movement. In some animals, the curled up position associated with increased sleep and decreased movement enables the animal to conserve heat. Decreased food intake in animals, specifically hunter-gatherers, minimizes heat loss associated with food seeking and allows the animal to conserve energy needed to facilitate a fever. According to this analysis, illness-induced anorexia will be most prevalent when an individual must work hard to obtain food (e.g., travel long distances from the safety of a home territory), and less prevalent when food is easily accessible. While Hart speculated that there were immediate benefits of decreased food intake in sick organisms, it should be acknowledged that in the long term illness-induced anorexia will become deleterious, because decreased food intake decreases the available energy stores.
According to Hart, these sickness behaviours are adaptive because they enhance illness-induced fever, which may aid the animal to combat pathogens. In fact, Kluger and colleagues have shown that animals not allowed to express the typical febrile response associated with illness have a greater risk of dying. Higher mortality rates were observed in lizards forced to remain in a cold environment after exposure to a bacterial infection than those placed in a warm environment . Moreover, lizards in the warm environment that were treated with an antipyretic (salicylate), experienced higher mortality rates than untreated animals . Likewise in humans, the use of antipyretics (aspirin and acetaminophen) has been shown to extend the duration and symptoms of chicken pox  and common colds L46], although we cannot be certain these effects of the drugs are attributable to the absence of a fever. The work of others has shown that a behavioural fever (moving to a warmer environment) increases survival, even if it increases the risk of predation, in animals such as flies, grasshoppers and fish (see ). It has been argued that the value of the febrile response derives from its ability to: enhance immune functioning (e.g., enhancing lymphocyte proliferation); make the body of the ill animal inhospitable for the pathogen by increasing body temperature above the optimal temperature for its survival and proliferation; and reduce iron concentrations, inhibiting the growth of many pathogens .
Following Hart's analysis, many people have investigated the behavioural changes induced by pathogens and immunologically activate agents. However, very little of this work has addressed Hart's main argument which was that behavioural change facilitates a fever in the sick animal and this facilitates its survival. While there is evidence to support the argument that fever is adaptive [43,44], this concept is quite controversial. Immunologically active agents do not necessarily induce fever, and not all fevers enhance survival, although interpretation of these data is not always straightforward [47,48], Importantly, sickness behaviour can be dissociated from fever. For example, McCarthy and colleagues have shown that administration of an antipyretic decreased endotoxin-induced fever without diminishing endotoxin-induced anorexia ,
Whether or not sickness behaviours are always expressed to enhance fever, there is evidence to support Hart's contention that sickness behaviours have survival value. For instance, mice that are not allowed to experience the typical anorexia associated with immune activation are less likely to survive endotoxin treatment. Murray and Murray infected mice with Listeria monocytogenes and allowed them to feed ad libitum, or they supplemented their food with gastric intubation , Intubated mice, fed daily to match uninfected controls, had lower survival rates than animals allowed to express Listeria-induced anorexia. Similarly, animals food-deprived for 24-72 hours prior to infection with Listeria monocytogenes survived at higher rates than animals that were not food-deprived . The deprived animals also had fewer bacteria in their spleen 2-4 days after infection. In a comprehensive analysis, Exton reviewed the evidence that anorexia is associated with enhanced immune functioning via increased activation of natural killer cells and macrophages, and increased T cell proliferation . Thus illness-induced anorexia may in part, facilitate survival because of an enhanced immune system response.
Most researchers that have systematically evaluated sickness behaviour have not measured mortality. Nevertheless, other support for Hart's proposal derives from work demonstrating that sickness behaviour is sensitive to test conditions, and does not necessarily occur in the same way for every animal, or every time an animal is sick. Moreover, immune system activation does not necessarily induce generalized decreases in behaviour. It has been shown that animals do not demonstrate typical sickness-induced behaviours when the consequences of doing so are detrimental [52,53] (see below). Variability in sickness behaviour has been suggested to reflect a differential sensitivity of animals to the perceived importance of a particular behaviour. The underlying assumption is that animals can choose to modify the induced sickness behaviour as a strategy to maximize its adaptive value, and it has been argued that animals are able to mitigate illness-induced behavioural changes so as to minimize the negative effects of sickness. This view assumes that animals evaluate their behaviour based on the possible consequences, which would be difficult to verify for nonhuman animals. Despite this, it does appear that the animals behave in a manner that facilitates their survival. Thus this view retains some appeal because, relying on evolutionary principles, it argues that these behavioural changes are adaptive for the animal.
Indeed, much of the variability in sickness behaviour appears to occur when the typical sickness behaviour profile could be detrimental to the organism's health. An example is a task in which rats or mice are allowed to investigate a multi-compartment chamber. The chamber is divided into nine compartments, each of which contains a loosely coiled ball mounted just below a round hole in the floor. Using this task, it has been shown that administration to mice of LPS, IL-la or IL-1 (3 all decreased exploration [38,54,55]. However, the effect depended upon the aspect of exploration assessed; compartment entries and nose pokes were much less sensitive to IL-1 and LPS than the times spent in contact with the novel stimuli. In other words, exploration of novel stimuli in the environment was more affected by the treatment than general exploration of the chamber. These diverse effects, decreased investigation of specific stimuli while continuing to investigate the entire chamber, can both be considered adaptive. Environmental investigation is important for an animal to ensure the safety of its environment, and thus, is to be expected during illness. On the other hand, investigation of the coiled ball prevents the animal paying attention to its global environment, and in times of weakness this could be dangerous. Therefore, decreased investigation of the novel stimuli may be adaptive for the animal. Because in this example, general motor activity is not impaired, it seems that motor activity is altered during sickness to the extent that it docs not impair the animal's safety.
Another example of the differential effects of immune activation is that food-restricted organisms are less likely to show illness- or cytokine-induced anorexia than non food-restricted organisms. Aubert and colleagues compared the effects of LPS on food hoarding and consumption in rats that received all their sustenance from hoarding with those in rats with supplemental food sources , They found that both groups of animals decreased food intake when they were ill. However, in the animals that obtained all their food from hoarding, LPS had less of an effect on hoarding as compared to animals with supplemental food sources. In other words, although LPS decreased eating at the time of analysis, it produced little disruption of hoarding for animals with no alternative food source , In another study, the effects of IL-1 (3 on sweetened milk intake were compared in food-restricted and free feeding mice , Mice were given the opportunity to consume sweetened milk in their home cage for 30 minutes. Milk intake in food-restricted mice was less affected by IL-1 (3 than intake in free feeding animals. Similarly, food-restricted mice trained to nose poke for milk using an operant paradigm were less affected by IL-1|3 than a group of free feeding mice also trained on the operant paradigm , Based on these data, it is reasoned that food-restricted animals have less stored energy and illness-induced anorexia and disruption of hoarding would be a greater threat so the reduced anorexia is a behavioural defence.
If illness-induced anorexia is health promoting because it promotes decreased activity and thus, conservation of heat, food intake with little behavioural cost should be less affected by immune system activation. Indeed, decreases in feeding-related behaviours are dependent upon how much energy must be expended to access the food. Using animals trained on operant schedules of food-maintained behaviour, it was shown that response cost modifies the IL-1-induced decreases in nose-poking for sweetened condensed milk . In these studies, mice were maintained on either a fixed ratio (FR) 4 (four responses necessary to obtain one drop of milk) or an FR32 (32 responses per drop) schedule of reinforcement. The behaviour of the animals on the FR32 schedule was affected by IL-1 (3 to a greater extent than animals on the FR4 schedule (i.e., a high response requirement is more readily affected by IL-1 (3 than a low response requirement) . Comparable results were obtained when the effect of LPS on animals maintained under an interval schedule (animals are reinforced with a food pellet for responding after a certain time has elapsed) was compared with those on a ratio schedule (animals are reinforced with a food pellet for responding a specified number of times). Interval schedules produce lower response rates and the cost of food is generally lower than ratio schedules. As expected, LPS had a more substantial effect on ratio than on interval schedules (Larson, unpublished observations). Furthermore, Larson and colleagues showed that IL-1 decreases behaviour maintained under progressive ratio schedules of responding . In these schedules, animals must produce an increasing number of responses in order to obtain the reinforcer (e.g., the first reinforcer may be earned after 10 responses; the second reinforcer after 20, the third after 30, etc.). These schedules require considerable behavioural output and, as expected, are sensitive to the effects of IL-1.
The above examples demonstrate that the profile of behavioural changes induced during illness is not stereotyped. In a healthy, well-fed, well-maintained organism, one might expect to observe all the behavioural responses described earlier. However, in an animal that has other constraints, such as a food-restriction, the profile of sickness behaviour may depend on these circumstances. Demonstrations of different effects of immune activation, depending upon circumstance, have been interpreted to suggest that immune activation affects motivation. Motivated behaviours can be defined as those associated with arousal, persistence and directional activity , Many data have shown that sickness decreases motivated behaviours, such as feeding, certain sexual activities, and behaviours maintained by pleasurable stimuli [e.g., 34,37, and 42], Thus it can be concluded that motivation to engage in certain behaviours is reduced during sickness, perhaps associated with immune activation. In instances where motivation for a particular behaviour is great (as in the case of food-restricted animals as discussed above), behaviours typically decreased by sickness (e.g., eating) are less affected. Thus as speculated by others [4,57], sickness may typically curtail behaviours that are unimportant in times of illness (e.g., sexual activity in females, social exploration). In cases where a particular behaviour might be important for survival (e.g., food intake in a food-restricted organism), an animal will be highly motivated to engage in the relevant behaviour, and the behaviour will be expressed.
Related to this, Aubert, Dantzer and colleagues [4,57] argued that sickness behaviours represent a motivational state, defining motivation as "a central state that organizes perception and action" [57, p. 1031]. Further, they suggested that other motivational states (e.g., fear, or in the case of the above example, hunger) may compete with sickness, and if other behaviours are more pressing, sickness behaviours may be superseded. Whether or not sickness should be regarded as a "motivational state", the evidence discussed above clearly indicates that sickness can affect motivation.
Illness appears to decrease many positively reinforced motivated behaviours but perhaps does so only to the extent that this is beneficial for the animal. Minimal work has evaluated the effect of illness on negatively reinforced motivated behaviours (i.e., behaviours motivated by the removal of aversive events). Miller reported that endotoxin did not disrupt operant behaviour reinforced by the elimination of a mildly aversive stimulus (a rotating drum) and actually increased it, even though the same dose of endotoxin disrupted food- and water-maintained behaviour . Another example is Gibertini's observations on the performance of rats in a Morris water-maze, which was impaired after Legionella administration . Interestingly, the magnitude of the disruption was dependent upon the water temperature in the water-maze; cooler water was associated with less impairment than was warmer water. To explain this, he suggested that the animals were more motivated to escape the cooler water and thus were less impaired by their illness . Taken together, these results suggest that behaviours followed by the removal of aversive stimuli are less likely to be affected by immune activators. Because maintaining behaviour that avoids aversive stimuli can in many cases be beneficial for the organism, these data further support suggestions that sickness behaviour is expressed in a manner that is affected by motivation and facilitates survival.
Another example of the dependence of sickness behaviour on the needs of the animal is evident in the effects of IL-1 on sexual behaviour. As indicated above, IL-1 markedly reduces sexual activity in female rats, but not in males . From a biological perspective, it would be detrimental for a sick female to add the burden of a foetus, the growth of which might also be adversely affected by the disease and its consequences. However, there are no such adverse consequences for the male. Indeed copulation may offer a last opportunity for the male to pass on its genes. These considerations may explain the lack of effect of IL-1 on sexual activity in male rats.
An issue rarely addressed by sickness behaviour researchers is the extent to which specific sickness behaviours might directly enhance survival (for example, if illness-induced anorexia were to facilitate an immune response), as opposed to enhancing survival simply by minimizing the risks associated with engaging in these behaviours. Infections and injuries produce effects that typically decrease most ongoing behaviours. Although this may occur to facilitate fever and to decrease to predators (see above), the decreased activity typically associated with sickness may also be adaptive because it minimizes further contact with additional pathogens, which could increase the animal's burden. Thus the reduction in ongoing activity may be the most important aspect of sickness behaviour. In cases where reduced activity would be deleterious (e.g., in food-restricted animals), sickness behaviour is less evident. Infections and injuries may elicit a nonspecific reduction in behaviour so that animals avoid high risk activities, and only perhaps, as Hart argued, enhance fever responses.
Conceiving of sickness behaviour as a nonspecific, general strategy that serves to minimize the activity of the organism does not imply that this profile occurs in the same way for every organism and/or in every circumstance. As described above, this profile may be affected by other factors (e.g., food restriction) that may attenuate typical sickness behaviours. However, the concept of sickness behaviour as a nonspecific, general defence helps us understand some aspects of sickness behaviour that do not readily appear adaptive, for example, the impaired cognitive function associated with immune activation [40,41]. It is not intuitively obvious why cognitive impairment would facilitate survival of the organism. However, impaired learning and memory may simply be a negative consequence of the global and nonspecific changes that serve to facilitate survival. Like other aspects of sickness behaviour, cognitive impairments are likely to be highly selective. According to this line of reasoning, sickness behaviours represent a general, evolutionarily advantageous defence that is adaptive for organisms facilitating their health. The profile is sensitive to environmental and physiological conditions imposed on the organism and
Was this article helpful?