Scientists Rediscover That Starvation Cures Type 2 Diabetes

Yes, you can reverse type 2 dia­betes if you starve your­self. In fact, a med­ical­ly super­vised water-only fast can be a use­ful way to man­age many dif­fer­ent kinds of diet-relat­ed dis­eases. For­tu­nate­ly, you do not have to starve your­self to reverse your type 2 dia­betes. Instead, you could sim­ply eat a low-fat, plant-based diet—like the pop­u­la­tions that don’t get type 2 dia­betes to begin with.

In June of 2011, some researchers from Britain pub­lished the results of a tri­al in which peo­ple with type 2 dia­betes who went on a star­va­tion diet (600 calo­ries per day) end­ed up with nor­mal fast­ing blood sug­ar lev­els. To me, that is not news. By 1841, a French phar­ma­cist named Apol­li­naire Bauchar­dat was rec­om­mend­ing that patients with what we now call type 2 dia­betes should eat as lit­tle as pos­si­ble and that they should fast occa­sion­al­ly to bring down their blood sug­ar. Since then, how­ev­er, dia­betes researchers have learned that it’s pos­si­ble to reverse type 2 dia­betes with­out such severe calo­rie restric­tion. In fact, I think that it’s bet­ter to teach peo­ple the diet that will enable them to cure their type 2 dia­betes with­in a cou­ple of weeks with­out lim­it­ing their food intake than to set them on a course of yo-yo diet­ing and pos­si­ble eat­ing dis­or­ders.

Bouchar­dat was one of the first clin­i­cians to put patients in charge of mon­i­tor­ing their own dia­betes. At first, his patients did this by keep­ing track of what they ate and tast­ing their urine to see how sweet it became. Lat­er, Bauchar­dat worked out a chem­i­cal test to detect sug­ar in urine. From mon­i­tor­ing the sug­ar con­tent of the urine, Bauchar­dat showed that when peo­ple with dia­betes ate sug­ars or starch­es, large amounts of sug­ar passed into their urine. The sug­ar in the urine reflect­ed high blood glu­cose lev­els. How­ev­er, the prob­lem in type 2 dia­betes is not that the per­son is eat­ing car­bo­hy­drates, it’s that the body has become resis­tant to the hor­mone insulin.

Start­ing in the 1930s, sci­en­tists start­ed to real­ize that fat­ty diets made the body less sen­si­tive to insulin, and that this insulin insen­si­tiv­i­ty was the under­ly­ing cause of the high blood sug­ar lev­els in peo­ple with type 2 dia­betes. Peo­ple who went on a low-fat, high-car­bo­hy­drate diet rapid­ly became more respon­sive to insulin.

Start­ing in the 1940s, Dr. Wal­ter Kemp­n­er at Duke Uni­ver­si­ty report­ed aston­ish­ing suc­cess in revers­ing type 2 dia­betes and dia­bet­ic com­pli­ca­tions with a diet based entire­ly on rice and fruit. Patients who found that they were los­ing too much weight on that low-fat diet were encour­aged to add pure white sug­ar to get more calo­ries. In Kempner’s report of 100 patients with dia­betes who were fed his high-car­bo­hy­drate, low-fat, low-pro­tein diet, most of the patients decreased their insulin dos­es and many dis­con­tin­ued tak­ing insulin. (It’s like­ly that some of the patients had type 1 dia­betes and there­fore would need to keep tak­ing insulin for the rest of their lives.)

The Amer­i­can Dia­betes Asso­ci­a­tion cur­rent­ly rec­om­mends that peo­ple with type 2 dia­betes eat lim­it­ed por­tions of foods from all of the four food groups. In 2006, how­ev­er, a clin­i­cal tri­al showed that the peo­ple who were ran­dom­ly assigned to eat as much as they liked of low-fat, unre­fined plant foods (75% car­bo­hy­drate by calo­rie) found it eas­i­er to stick to their diet, lost more weight, and made faster progress in revers­ing their dia­betes than did the peo­ple who were ran­dom­ly assigned to fol­low the ADA’s rec­om­men­da­tions.

Behind Barbed Wire_PrintNote: In my book Thin Dia­betes, Fat Dia­betes: Pre­vent Type 1, Cure Type 2, I explain the rela­tion­ship between body weight and blood sug­ar. French doc­tors have always used the term fat dia­betes (dia­bètes mai­gre) to refer to the rel­a­tive mild form of dia­betes that occurs in peo­ple who are at least a lit­tle bit over­weight and that goes away if they lose weight. Fat dia­betes is the body’s way to avoid stor­ing too much of the fat from a fat­ty diet. If you have fat dia­betes, it means that you are a nat­u­ral­ly thin per­son. It means that your body is will­ing to sac­ri­fice everything—your feet, your eye­sight, your kid­neys, and even your life—to keep you from gain­ing any more weight. The solu­tion to this prob­lem is to switch to a low-fat, high-car­bo­hy­drate, high-fiber diet. This diet revers­es type 2 dia­betes and is also good for peo­ple with thin dia­betes (type 1 dia­betes).

Only a Small Percentage of Diabetes Cases Are Truly Genetic

Late­ly, genes have been all the rage in med­ical research. Many peo­ple hope that genet­ic stud­ies will unrav­el the secrets of many dis­eases that, strict­ly speak­ing, aren’t real­ly genet­ic. Some­times, peo­ple get so caught up in think­ing about the genes that could be involved in a dis­ease that they ignore the impor­tance of diet to pre­vent and man­age the dis­ease. Yet even for a dis­ease that is tru­ly 100% genet­ic, dietary man­age­ment can be impor­tant. The clas­sic exam­ples are the rare forms of dia­betes mel­li­tus that real­ly are genet­ic. These forms of dia­betes are called mono­genic dia­betes mel­li­tus. Some­times they’re called MODY, which stands for matu­ri­ty-onset dia­betes of the young. It is a form of mono­genic dia­betes, which means that it is due to a muta­tion in a sin­gle gene.

There are sev­er­al dif­fer­ent forms of MODY, each result­ing from the muta­tion of a dif­fer­ent gene. These dis­or­ders tend to be auto­so­mal dom­i­nant, which means that if you inher­it the gene from one of your par­ents, you have the dis­ease. Clin­i­cal­ly, cas­es of MODY tend to look like a hybrid between type 1 and type 2 dia­betes. Togeth­er they account for some­where between 1% and 5% of cas­es of dia­betes mel­li­tus.

Dia­betes mel­li­tus refers to sev­er­al unre­lat­ed con­di­tions that all result in high lev­els of sug­ar in the blood­stream. In the 1870s, French researchers noticed that there are two main types of dia­betes mel­li­tus. One is a cat­a­stroph­ic dis­ease that tends to strike chil­dren and young adults. Before the dis­cov­ery of insulin, these patients would rapid­ly lose weight, go into a coma, and die with­in a mat­ter of days or weeks, regard­less of treat­ment. The French researchers called it dia­bète mai­gre, or thin dia­betes. Eng­lish-speak­ers called it juve­nile dia­betes, then insulin-depen­dent dia­betes, and now type 1 dia­betes. In con­trast, the most com­mon form of dia­betes is a much milder, reversible con­di­tion that occurs in mid­dle-aged to elder­ly adults who are usu­al­ly at least a bit over­weight. The French called this con­di­tion dia­bète gras, or fat dia­betes. Eng­lish-speak­ers tact­ful­ly referred to this kind of dia­betes as adult-onset, then non–insulin-dependent, and now type 2 dia­betes.


If left untreat­ed, all forms of dia­betes mel­li­tus result in abnor­mal­ly high lev­els of a sug­ar called glu­cose in the blood­stream. In cas­es of type 1 dia­betes, the high glu­cose lev­els result from the fail­ure of the pan­creas to pro­duce the hor­mone insulin. In cas­es of severe insulin short­age, the body goes into a state called dia­bet­ic ketoaci­do­sis, which is a med­ical emer­gency.

Ear­ly on, doc­tors noticed that peo­ple with type 2 dia­betes rarely get dia­bet­ic ketoaci­do­sis. In severe cas­es, they might go into a coma because their blood sug­ar is so high that they get severe­ly dehy­drat­ed, but they rarely get such severe keto­sis that their blood pH is affect­ed. That’s because peo­ple with type 2 dia­betes are pro­duc­ing some insulin. In fact, they tend to have abnor­mal­ly high lev­els of insulin in their blood­stream. The prob­lem is that their cells have become resis­tant to the effects of insulin.

MODY was dis­cov­ered in the 1960s, when physi­cians noticed that some chil­dren had per­sis­tent, symp­tom­less high blood sug­ar that didn’t progress to ordi­nary type 1 dia­betes. In oth­er words, the con­di­tion looked like type 2 dia­betes, except that it was hap­pen­ing in young peo­ple who weren’t over­weight. The child’s pan­creas was pro­duc­ing some insulin, just not enough to meet the body’s needs. By the 1970s, it was clear that this prob­lem ran in fam­i­lies. At present, at least 10 forms of MODY have been described, each of them linked to a dif­fer­ent gene.

The treat­ment for MODY depends on what type of MODY the patient has. Genet­ic test­ing is now avail­able for MODY1, MODY2, MODY3, MODY4, and MODY5. MODY2 is a rel­a­tive­ly mild dis­ease that can be man­aged by diet and exer­cise. As in ordi­nary cas­es of type 2 dia­betes, the goal is to avoid or reverse insulin resis­tance, so that the amount of insulin the pan­creas pro­duces on its own will be enough to reg­u­late the patient’s blood sug­ar nat­u­ral­ly. Main­tain­ing insulin sen­si­tiv­i­ty through prop­er diet and exer­cise is undoubt­ed­ly also impor­tant in man­ag­ing the oth­er, more severe forms of MODY. Patients with MODY1, MODY3, or MODY4 are usu­al­ly giv­en a sul­fony­lurea drug that caus­es the pan­creas to secrete extra insulin. MODY5 is a severe con­di­tion that is usu­al­ly treat­ed as if the patient has no pan­creas what­so­ev­er; the patient is giv­en insulin injec­tions as well as diges­tive enzymes that are tak­en by mouth.

Even­tu­al­ly, doc­tors found that a few new­borns had a con­di­tion that seemed like type 1 dia­betes. This con­di­tion, called neona­tal dia­betes, is gen­er­al­ly inher­it­ed as an auto­so­mal reces­sive trait. To get a case of infan­tile dia­betes, the baby has to inher­it a defec­tive ver­sion of a gene from both par­ents.

If you think that you or some­one in your fam­i­ly has either MODY or neona­tal dia­betes, con­tact the Nation­al Cen­ter for Mono­genic and Neona­tal Dia­betes at the Uni­ver­si­ty of Chica­go.

Behind Barbed Wire_PrintNote: For more infor­ma­tion about dia­betes in gen­er­al and the genet­ic forms of dia­betes in par­tic­u­lar, see my book Thin Dia­betes, Fat Dia­betes: Pre­vent Type 1, Cure Type 2.

We’ve Known Since the 1930s: Fatty Diets Cause Insulin Resistance

This graph shows the glu­cose tol­er­ance test results of a healthy per­son on a high-fat test diet as opposed to a high-carb diet. Notice that he had low­er, more sta­ble blood sug­ar lev­els after eat­ing a high-carb diet!

Low-carb gurus have been telling peo­ple that insulin resis­tance, which is the under­ly­ing cause of type 2 dia­betes, results from eat­ing too much car­bo­hy­drate. Yet sci­en­tists have known since the 1930s that the prob­lem can be pro­voked by a high-fat diet and reversed by a starchy, low-fat diet. You can read the research for your­self here.

This work was done in the 1930s. How long will it take before the low-carb gurus hear about it?