Is a single measurement of Lp(a) sufficient for risk determination?