I really dislike the unit test and integration test analogy. It is based on the assumption that the human body has something in common with a computer program - or anything man-made for that matter. This is clearly not the case. We tend to think of ourselves in terms of whatever our current lead technology is. It has been like that before [1]. While it is certainly psychologically comforting it becomes dangerous when we act based on what really is somewhat of an arbitrary model.
It's based on the assumption that the human body is a complex organism made up of interacting subsystems, and that there is a distinction between testing the effect of a change on one subsystem and testing the effect of a change on the entire organism. That's a very abstract assumption, and easily applies in this case. If you can't see anything in common between the human body and a computer program, you're just not looking at an abstract enough level. You can start with "they both exist" and work your way towards the concrete from there.
> We tend to think of ourselves in terms of whatever our current lead technology is. It has been like that before [1].
This is what we call progress. Technology doesn't grow on its own, it embodies the deep insights into workings of reality humans gain throughout the century. Therefore it is to be expected that we will draw parallels and that they will improve as we improve our tech.
[1]: http://webadventures.at/2007/06/17/der-mensch-als-industriep...