Transdisciplinary (TD) approaches are increasingly used to address complex public health problems such as childhood obesity. Compared to traditional grant-funded scientific projects among established scientists, those designed around a TD, team-based approach yielded greater publication output after three to five years. However, little is known about how a TD focus throughout graduate school training may affect students' publication-related productivity, impact, and collaboration. The objective of this study was to compare the publication patterns of students in traditional versus TD doctoral training programs. Productivity, impact, and collaboration of peer-reviewed publications were compared between traditional (n = 25) and TD (n = 11) students during the first five years of the TD program. Statistical differences were determined by t-test or chi square test at p < 0.05. The publication rate for TD students was 5.2 ± 10.1 (n = 56) compared to 3.6 ± 4.5 per traditional student (n = 82). Publication impact indicators were significantly higher for TD students vs. traditional students: 5.7 times more citations in Google Scholar, 6.1 times more citations in Scopus, 1.3 times higher journal impact factors, and a 1.4 times higher journal h-index. Collaboration indicators showed that publications by TD students had significantly more co-authors (1.3 times), and significantly more disciplines represented among co-authors (1.3 times), but not significantly more organizations represented per publication compared to traditional students. In conclusion, compared to doctoral students in traditional programs, TD students published works that were accepted into higher impact journals, were more frequently cited, and had more cross-disciplinary collaborations.