Biomedical Engineering Reference
In-Depth Information
reactions to disease by both sick and uninfected individuals, and heterogeneous
basal host resistance that varies as a function of social contact. The observed
disease trajectories show knife-edge dynamics, with epidemics burning slowly
through a population for a variable period of time before either collapsing or
exploding. These "time-bomb" kinetics are critical to recognize, but they would
not be apparent from the "expected-value" disease trajectories generated by tra-
ditional algebraic models (e.g., the falsely consoling prediction limits of Figure
7E). The qualitative significance of variability is also apparent in the contrast
between Figure 7A and 7C. Mean trajectories are comparable, so conventional
epidemiologic models would suggest little difference. However, host popula-
tions consistently survive in panel A whereas large segments of society are often
annihilated in C. The significance of that difference transcends public health to
reach the level of evolutionary extinction.
Throughout their evolutionary histories, vertebrates and their parasites have
each shaped the other's development (1,14,17,18). The present studies suggest
that a similar reciprocal dynamic may have occurred in the evolution of the im-
mune and nervous systems. As biochemical crosstalk between these two systems
becomes increasingly appreciated, one teleologic perspective has emerged to
suggest the immune system inhibits social behavior to maximize its own claim
on physiologic resources (3,8). The present analyses support an alternative view
in which biologically induced sickness behavior generates an emergent social
immune response that operates in synergy with leukocytes to defend its genome
at a species-wide level. From this perspective, the jaggedly unpredictable dis-
ease trajectories seen in many of these studies testify to the close fight between
socially defended hosts and their socially predatory pathogens. Both the immune
system and the nervous system have evolved under the weight of this pressure,
and the present results suggest that they are more likely to collaborate than com-
pete in response.
5.
ACKNOWLEDGMENTS
This work was supported by the National Institutes of Allergy and Infec-
tious Disease (AI49135, AI52737), the James L. Pendelton Charitable Trust, and
a visiting scholarship from the Santa Fe Institute for Complex Systems.
6.
APPENDIX
This appendix provides more detail on the implementation of ActiveHost as
summarized in Figure 1. This agent-based modeling system is composed of four
basic objects: (a) a
SimulationSystem
that creates multiple instances of a given
