The components model of addiction posits that all addictions share six components, namely salience, tolerance, mood modification, relapse, withdrawal, and conflict. This highly influential model has resulted in the development of numerous psychometric instruments that measure addictive behaviors according to these criteria. However, recent research suggests that, in the context of behavioral addictions, certain components constitute peripheral features that do not distinguish non-pathological from pathological behavior. Using "addictive" use of social media as a representative example, we examined this perspective by testing whether these six components actually assess central features of addiction, or whether some of them constitute peripheral features that are not indicative of a disorder. Four independent samples totaling 4,256 participants from the general population completed the Bergen Social Media Addiction Scale, a six-item psychometric instrument derived from the components model of addiction to assess social media "addiction". By performing structural equation modeling and network analyses, we showed that the six components did not form a unitary construct and, crucially, that some components (i.e., salience, tolerance) were not associated with measures assessing psychopathological symptoms. Taken together, these results suggest that psychometric instruments based on the components model conflate central and peripheral features of addiction when applied to behavioral addictions. This implies that such instruments pathologize involvement in appetitive behaviors. Our findings thus call for renewing the conceptualization and assessment of behavioral addictions.