Sigma terms and strangeness content of the nucleon with Nf=2+1+1 twisted mass fermions

We investigate excited state contaminations in a direct computation of the nucleon $\sigma$-terms. This is an important source of systematic effects that needs to be controlled besides the light quark mass dependence and lattice artefacts. We use maximally twisted mass fermions with dynamical light ($u$,$d$), strange and charm degrees of freedom. Employing an efficient stochastic evaluation of the disconnected contribution available for twisted mass fermions, we show that the effect of excited states is large in particular for the strange $\sigma$-terms, where it can be as big as $O(\gtrsim 40%$). This leads to the unfortunate conclusion that even with a source-sink separation of $\sim 1.5 \fm$ and a good statistical accuracy it is not clear, whether excited state effects are under control for this quantity.