Let's look at what the submitted article, which is a press release from a university, says. "In both mice and a Phase 1 human clinical trial, long periods of not eating significantly lowered white blood cell counts. In mice, fasting cycles then 'flipped a regenerative switch,' changing the signaling pathways for hematopoietic stem cells, which are responsible for the generation of blood and immune systems, the research showed."

What's good about this line of research is that it is trying to show whether or not an observation found in a model animal will also be found in human patients. As preliminary research, this will have to be replicated by other researchers before we can rely on this finding (NOT extended fully to human patients in the current study) to guide treatment of human patients.

This is good news that people are investigating this issue. Once there has been a thorough review article on this issue published in a different journal by a different author, summing up several well designed studies, then we will really have something to talk about here on Hacker News.

I should comment on some of the other comments here. One very tricky problem in studies of human nutrition is that there isn't a good model organism for nutrition and its effect on human health. The study reported here, just like most medical intervention studies, begins in a mouse model. Mice are well understood organisms and their similarities to and differences from human beings for many medical treatments are well understood. Mice are not a particularly good model for nutrition studies, however, because mice are rodents (part of a clade of obligate herbivores) while human beings are primates (part of a clade of facultative omnivores). Moreover, human beings, the current species Homo sapiens, have evolved with co-evolution of the gut in the environment of the cultural practice of cooking food.[1] Cooking is a human cultural universal. No other animal lineage has evolved in a similar environment, so no animal provides a fully suitable model for studies of human nutritional interventions.

Human nutrition studies are HARD, because they require minute-by-minute monitoring of the subjects and exact measurements of food intake to gather meaningful experimental data. (I recall a TV news report from the 1970s about a human nutrition study in which the study volunteers, who of course were paid for this, lived confined inside a lab in which lab technicians weighed all their food to the nearest gram and controlled everything they could eat for the duration of the experiment. Alas, I've never heard of results of that study, perhaps because the sample size, with such an expensive procedure, was too small to generate meaningful data.) Yes, let's see what nutritional interventions do what for human beings, but let's be careful not to jump to conclusions too soon, because careful data gathering on this topic is especially difficult, and anecdotes crowd out data in most popular discussion of this topic.

[1] http://www.scientificamerican.com/article/cooking-up-bigger-...

http://www.popsci.com/science/article/2012-10/eating-cooked-...

http://www.smithsonianmag.com/science-nature/why-fire-makes-...