Interestingly, the glucose idea is one that is often thrown around with "ego depletion", but there are several problems with it. If the replicability issues are set aside, there are many articles which suggest people's beliefs about "ego depletion" change the observed effects (e.g. [1]). More importantly, there's good reason to believe that early studies in support of the glucose hypothesis did not make sense [2].

1: http://pss.sagepub.com/lookup/doi/10.1177/0956797610384745

2: Slate article, but mentions some key studies, http://www.slate.com/articles/health_and_science/cover_story...