Complex networks can exhibit different behaviors, there can be networks where the nodes, after being created, can continue to increase or decrease the number of their incoming and/or outgoing links. An example of these networks is the WWW, where the webmaster can add or delete hyperlinks on the web pages under their responsibility at any time. There are also networks where this can not happen. For example, in citation networks of scientific articles, after an article has been published, it will start gaining incoming links as it is cited, but the outgoing links will remain unchanged. Although articles are published with a fixed number of references, the distribution of their outgoing links follows a power law, as if they were the result of a preferential process. So how can we explain that the number of references an author includes in a scientific article is not purely random? In this work, a growth model of this type of complex network is presented, where it is proposed that the distribution of outgoing links can be shaped by the presence of communities.