How independent are "independent" effects? Relative risk estimation when correlated exposures are measured imprecisely.

How independent are "independent" effects? Relative risk estimation when correlated exposures are measured imprecisely.