Biomedical Engineering Reference
In-Depth Information
sense that individual lifespan is shortened by the square of its intensity (.5 in-
flammation results in a 25% reduction in average lifespan) (15). Sickness behav-
ior is modeled as a multiplicative link between the magnitude of the inflamma-
tory response and the fractional reduction in social contacts realized during
inflammation (sickness behavior sensitivity parameter = 1.0 initially for all indi-
viduals, with the parental sensitivity value passed on to progeny with 5% noise,
as described above for the inflammatory response). Each individual produces 2
progeny at random times between 13 and 40 years of age and, in the absence of
infection, dies of natural causes at a normally distributed age with a mean of 40
and standard deviation of 10. The population mean value of the inflammatory
response parameter and the sickness behavior sensitivity parameter are averaged
over 20 simulations of 250 time units (~10 generations). The gray line in Figure
8 reveals a strong selective pressure for increased inflammatory response that
begins to decelerate at ~40% resistance as the costs of septic shock, autoimmu-
nity, and infertility begin to offset protection from infectious disease. In contrast,
the sickness behavior sensitivity parameter (black line) shows slow initial
growth that subsequently accelerates as inflammatory responses become pro-
nounced (the log of the mean sickness gain parameter is plotted for comparison
with the linear inflammatory parameter). During the early evolution of the im-
mune system (0-100 time units), there is little selective pressure to link social
behavior with the more directly effective inflammatory responses. The sickness
behavior sensitivity parameter multiples the effects of biological inflammation,
so it fails to evolve much while the basic inflammatory response is weak. How-
ever, once inflammatory responses begin to reach their cost-induced limits (15),
benefits begin to accrue to those who reduce social contact in response to in-
flammation. Interestingly, in models that suppress the emergence of sickness
behavior, the biological immune response evolves more rapidly and reaches a
higher asymptotic equilibrium. This suggests that behavioral immune response
contributes significantly to total host protection. Such results also imply that we
may have been spared higher rates of inflammatory disease by the emergence of
CNS-mediated sickness behavior. This is certainly consistent with the observa-
tion that vertebrate physiology dedicates substantial molecular resources to
communication between the disease-sensing immune system and the behavior-
controlling nervous system (2,16,17). Similar selective pressures may also have
shaped mammalian brains to prefer small clustered social systems rather than
large herds. Neither structural sparsity or disease-reactive linkage slows an epi-
demic much in isolation, but their combination can be decisive. In the model of
Figure 7, for example, the joint effects of a clustered social structure and dis-
ease-reactive connectivity are equivalent to a 60-fold increase in the strength of
biological immunity. The sick are unlikely to take much consolation from the
fact that their immune systems synthesize most of their suffering to protect oth-
ers, but they may find some comfort in considering the more malevolent blood
they have been spared.
