This paper addresses the problem of parameter dependent state feedback control (i.e. gain scheduling) for linear systems with parameters that are assumed to be available (measured or estimated) in real time and are allowed to vary in a compact polytopic set with bounded variation rates. A new sufficient condition given in terms of linear matrix inequalities permits to determine the controller gain as an analytical function of the time-varying parameters and of a set of constant matrices. The closed-loop stability is assured by means of a parameter dependent Lyapunov function. The condition proposed encompasses the well-known quadratic stabilizability condition and allows to impose structural constraints such as decentralization to the feedback gains. Numerical examples illustrate the efficiency of the technique.