redundancy in the Friedmann equations

Under which circumstances is the \(\dot{a}\)-Friedmann equation the integral of the \(\ddot{a}\)-equation? What acts as the integration constant? Why would that particular variable not be able to introduce de/ac-celeration?

bonus question: Why is (in contrast to real analysis) the derivative of a complex differentiable function always again complex differentiable?