Age matching animal models to humans--theoretical considerations.

Biomedical animal models predict clinical efficacy with varying degrees of success. An important feature of in vivo modeling is matching the age of the animals used in preclinical research to the age of peak incidence for a disease state in humans. However, growth and development are highly variable between mammalian species, and age matching is always based on assumptions about the nature of development. We propose that researchers commonly make the assumption that developmental sequences are highly conserved between mammalian species--an assumption that we argue is often incorrect. We instead argue that development is often a modular process. Consideration of the modular nature of development highlights the difficulty in matching animal ages to human ages in a one-to-one scalar manner. We illustrate this with a discussion of the problem of age matching rodents to humans for neuroprotection experiments, and argue that researchers should pay deliberate attention to the modularity of developmental processes in order to optimally match ages between species in biomedical research.