In 1908, Linda Hazzard, an American with some training as a nurse, published "Fasting for the Cure of Disease," which claimed that minimal food was the route to recovery from a variety of illnesses, including cancer. Hazzard was jailed after one of her patients died of starvation. But what if she was, at least partly, right?

A new surge of interest in fasting suggests that it might indeed help people with cancer. It might also reduce the risk of developing cancer, guard against diabetes and heart disease, help control asthma and even stave off Parkinson's disease and dementia.

"We know from animal models," says Mark Mattson at the National Institute on Aging, "that if we start an intermittent fasting diet at what would be the equivalent of middle age in people, we can delay the onset of Alzheimer's and Parkinson's."

Until recently, most studies linking diet with health and longevity focused on calorie restriction. They have had some impressive results, with the life span of various lab animals lengthened by up to 50 percent after their caloric intake was cut in half. But these effects do not seem to extend to primates. A 23-year study of macaques found that although calorie restriction delayed the onset of age-related diseases, it had no impact on life span. So other factors, such as genetics, may be more important for human longevity.

That's bad news for anyone who has gone hungry for decades in the hope of living longer, but the finding has not deterred researchers who study fasting. They point out that although fasting obviously involves cutting calories — at least on specific days — it brings about biochemical and physiological changes that daily dieting does not. Besides, calorie restriction may leave people susceptible to infections and biological stress, whereas fasting, done properly, should not.

Some even argue that we are evolutionarily adapted to going without food intermittently. "The evidence is pretty strong that our ancestors did not eat three meals a day plus snacks," Mattson says. "Our genes are geared to being able to cope with periods of no food."

---

Trying out a fast

Fasting will leave you feeling crummy in the short term because it takes time for your body to break psychological and biological habits, researchers say. There isn't really agreement, though, on what fasting entails. To research this article, I am trying out the "5:2" diet, which allows me 600 calories in a single meal on each of two weekly "fast" days. (The normal recommended daily intake is about 2,000 calories for a woman and 2,500 for a man.) Proving that fasting is not necessarily about losing weight, I am allowed to eat whatever I want on the five non-fast days.

A more draconian regimen than the 5:2 plan has restricted-calorie fasts every other day. Then there's total fasting, in which participants go without food for one to five days. (Fasting for more than about a week is considered dangerous.) This might be a one-off experience, or repeated weekly or monthly.

Different regimens have different effects on the body. A fast is considered to start about 10 to 12 hours after a meal, when you have used up all the available glucose in your blood and start converting glycogen stored in liver and muscle cells into glucose to use for energy. If the fast continues, there is a gradual move toward breaking down stored body fat, and the liver produces "ketone bodies," short molecules that are byproducts of the breakdown of fatty acids. These can be used by the brain as fuel. This process is in full swing three to four days into a fast.

Various hormones are also affected. For example, production of insulin-like growth factor 1 (IGF-1) drops early and reaches very low levels by Day 3 or 4. It is similar in structure to insulin, which also becomes scarcer with fasting, and high levels of both have been linked to cancer.

As for treating cancer, Valter Longo, director of the Longevity Institute at the University of Southern California, thinks that short-term complete fasts maximize the benefits. He has found that a 48-hour total fast slowed the growth of five of eight types of cancer in mice, the effect tending to be more pronounced the more fasts the animals undertook.

Fasting is harder on cancer cells than on normal cells, he says. That's because the mutations that cause cancer lead to rapid growth under the physiological conditions in which they arose, but they can be at a disadvantage when conditions changes. This could also explain why fasting combined with conventional cancer treatment provides a double whammy. Mice with gliomas — very aggressive forms of cancer and the most commonly diagnosed brain tumor in people — were more than twice as likely to survive a 28-day study if they underwent a 48-hour fast accompanied by radiation therapy as were those that did not fast.

Could fasting prevent cancers from developing in the first place? Evidence is scant.

Longo says there are "very good reasons" why it should. He points out that high levels of IGF-1 and glucose in the blood and being overweight are risk factors for cancer, and they can all be improved by fasting.

Another risk factor is insulin, says Michelle Harvie at Britain's University of Manchester. Studying a group of women whose family history put them at high risk of developing breast cancer, she put half of them on a diet that cut calories by about 25 percent and half on a 5:2 fast. After six months, both groups showed a reduction in blood insulin levels, but the reduction was greater in the fasting group. Harvie's team is now analyzing breast biopsies to see whether this translates to fewer of the genetic changes associated with increased cancer risk.

---

The effect on diabetes

High insulin is also associated with Type 2 diabetes, so perhaps it is no surprise that fasting shows promise there, too. At the Intermountain Heart Institute in Murray, Utah, Benjamin Horne has found that a 24-hour water-only fast, performed monthly, raises levels of human growth hormone. That hormone triggers the breakdown of fat for energy use, reducing insulin levels and other metabolic markers of glucose metabolism. As a result, people lost weight, and their risk of getting diabetes and coronary heart disease was reduced. Alternate-day fasting (with a 500-calorie lunch for women and a 600-calorie meal for men on fast days) has similar benefits, says Krista Varady of the University of Illinois. She has seen improvements in people's levels of low-density lipoprotein cholesterol, or "bad cholesterol," and blood pressure in volunteers eating either a low-fat or high-fat diet on "feeding" days.

For people who are overweight, any kind of intermittent fasting diet will probably help reduce the risk of diabetes and cardiovascular problems, Mattson says. In 2007, he found another benefit, too. He put 10 overweight people with asthma on an alternate-day incomplete fast and found that their asthma symptoms improved after just a few weeks. Blood markers of inflammation, including C-reactive protein, also decreased, suggesting that the fast was helping to moderate their overactive immune system.

Whether fasting would benefit normal-weight people with asthma or other conditions associated with an overactive immune response remains to be seen. There is some evidence that alternate-day fasting can lower their levels of blood fat. However, Mattson suspects that with diabetes and cardiovascular disease, fasting may not be as beneficial for people of normal weight as it is for people who are overweight, simply because they are already likely to be in pretty good shape, metabolically speaking.

---

How the brain reacts

Mattson has, however, identified another effect of fasting that he believes can benefit everyone: It is good for the brain. "If you look at an animal that's gone without food for an entire day, it becomes more active," he says. "Fasting is a mild stressor that motivates the animal to increase activity in the brain." From an evolutionary perspective, this makes sense, because if you are deprived of food, your brain needs to work harder to help you find something to eat.

His studies suggest that alternate-day fasting, with a single meal of about 600 calories on the fast day, can boost the production of a protein called brain-derived neurotrophic factor by 50 to 400 percent, depending on the brain region. This protein is involved in the generation of new brain cells and plays a role in learning and memory. It can also protect brain cells from the changes associated with Alzheimer's and Parkinson's. In mice engineered to develop Alzheimer's-like symptoms, alternate-day fasting begun in middle age delayed the onset of memory problems by about six months. "This is a large effect," Mattson says, perhaps equivalent to 20 years in humans.

So, what about the common advice to start the day with a good breakfast? Mattson believes it is flawed, pointing out that the studies supporting this idea were based on schoolchildren who usually ate breakfast; a decline in their academic performance might simply be due to the ill effects that occur when people begin fasting.

Mattson skips breakfast and lunch five days a week, then has dinner and normal weekend meals with his family. Varady has tried alternate-day fasting, but she likes to eat dinner with her 18-month-old child and husband, so she does all her eating within an eight-hour period each day.

Harvie sounds a cautious note for anyone thinking of giving fasting a go. "We still don't know exactly who should be fasting, how often or how many days a week," she says. Also, it may not be without risks. One study in rats, for example, found that an alternate-day fast for six months reduced the heart's ability to pump blood.

There is also the fact that fasting is difficult. Varady finds that 10 to 20 percent of people who enroll in her studies drop out, unable to stick to the regime. This may be less of a problem in the future, though. Some researchers are investigating the possibility that you can get some of the health benefits of fasting simply by reducing protein intake.

As I count down the minutes to the end of my fast, I can't help but wish them success.

---

This story was produced by New Scientist magazine and can be read in full at www.newscientist.com.