Ketogenic diets (KDs) have gained popularity among patients and researchers alike due to their putative anti-tumor mechanisms. However, the question remains which conclusions can be drawn from the available human data thus far concerning the safety and efficacy of KDs for cancer patients. A realist review utilizing a matrix analytical approach was conducted according to the RAMESES publication standards. All available human studies were systematically analyzed and supplemented with results from animal studies. Evidence and confirmation were treated as separate concepts. In total, 29 animal and 24 human studies were included in the analysis. The majority of animal studies (72%) yielded evidence for an anti-tumor effect of KDs. Evidential support for such effects in humans was weak and limited to individual cases, but a probabilistic argument shows that the available data strengthen the belief in the anti-tumor effect hypothesis at least for some individuals. Evidence for pro-tumor effects was lacking completely. Feasibility of KDs for cancer patients has been shown in various contexts. The probability of achieving an anti-tumor effect seems greater than that of causing serious side effects when offering KDs to cancer patients. Future controlled trials would provide stronger evidence for or against the anti-tumor effect hypothesis.
Keywords: Cancer; Evidence; Ketogenic diet; Realist review; Realist synthesis.