Visible Light Communication (VLC) is gaining interest as a viable option to supplement next generation wireless communication systems with dense directional wireless cells. The effect of interference is important when analyzing aggregate performance of these systems. Just as signal to noise ratio (SNR) is used to evaluate link performance when disturbances are primarily related to noise, signal to interference ratio (SIR) and signal to interference plus noise ratio (SINR) are metrics of interest for evaluation of systems in the presence of interference. For these metrics to be used interchangeably with SNR, interference should be Normally distributed. This assumption is valid in the presence of many interferers when the central limit theorem applies; however, directionality in VLC systems often leads to scenarios with dominant interferers. In this paper, we analyze the validity of the Gaussian assumption in VLC systems and show that the theoretical error rate can be distorted by 100× or more when the assumption is erroneously observed. In addition, we define the distribution of commonly used VLC modulation schemes in relation to average optical power in order to provide fair comparison under constraints of the optical channel.