Does fasting or intermittent fasting help prevent cancer?
There is no evidence that fasting prevents cancer in humans; the animal results are promising but contradictory. If you want to address excess weight as a way of reducing risk, intermittent fasting can be a tool for that, but do not expect a direct cancer-preventive effect.
There are no randomised trials in humans showing that intermittent fasting or a fasting-mimicking diet (FMD) reduces the risk of cancer. However promising the theory may sound, the effects on cancer incidence in humans are currently unknown.
In animal research the picture looks more encouraging. Mouse and rat studies show protective effects against several types of cancer. These effects are attributed to reductions in insulin, growth-hormone-related signalling and leptin. However, in some animal models harmful effects were also found, depending on the type of cancer. Animal research is therefore far from clear-cut, and in any case the step from mouse to human is a large one.
An interesting line of research combines FMD with hormone therapy in breast cancer. In mouse models, that combination led to prolonged tumour suppression. In a small group of patients with hormone-receptor-positive breast cancer, comparable metabolic changes were measured, such as lower insulin and leptin levels. Whether this also translates into clinical benefit still needs to be investigated in larger studies.
Fasting may contribute indirectly to cancer prevention through weight management, as excess weight is a well-established risk factor for cancer. Whether fasting itself, independent of weight loss, changes any cancer-related processes in the body has not yet been demonstrated in humans.
For people already undergoing cancer treatment: early clinical studies suggest that cyclical fasting is generally safe and may reduce chemotherapy side effects by protecting normal cells. However, this evidence is still too preliminary to act on. Outside of a clinical trial, fasting during active cancer treatment is not currently recommended.
All claims are based on PMID 34383300, 38147966, 39059383, 32669709, 30327499, 24440038, 27810402, 39064705. No completed randomised trials in humans are available that measure cancer incidence as an outcome.