Cathy entered the weight-loss program at the University of Pennsylvania Medical School with a warm personality, an infectious smile and a great desire to be thin.
My colleagues and I were beginning to experiment with a very low-calorie diet at our clinic, a fast combined with nutrition supplements that was designed to get rid of a lot of weight very quickly. The diet had just 420 calories per day.
Cathy did well on it initially, dropping from 230 to 193 pounds. Then she stopped losing weight. Something was wrong. A 193-pound person should certainly lose weight on 420 calories a day. Was she cheating on her diet? We did not believe so. Would more exercise help? She was already walking up to 3 miles a day.
Cathy was heartbroken. She wanted desperately to slim down: She felt it was important for her energy level, attractiveness, self-image and health. We wanted to help, but she was already on the strictest diet at our disposal. I began to wonder why Cathy's body was resisting weight loss. How could her body be so efficient, maintaining so much weight on so few calories?
Like most people in our clinic, Cathy had lost weight in the past, but she had regained it. My impression was that patients who had dieted many times in their lives had the most trouble losing weight in our program. At first glance, this seemed logical. Veteran dieters probably had some physiological disadvantage, such as a low metabolic rate--the rate at which their bodies use up energy from food and other sources. This made them gain weight in the first place, I thought, and then made it difficult for them to lose weight and keep it off.
Cathy, however, prompted me to take a different perspective. Was it possible that previous diets actually created this tendency to maintain more than 190 pounds on just a few calories? Could dieting inhibit later weight loss?
These questions generated a project now involving researchers from five universities who are focusing on "yo-yo dieting": repeated cycles of weight loss and gain. Many people diet this way, even those who are not overweight. We are examining the effect of yo-yo dieting on metabolism and health.
This dieting pattern stems from the tremendous cultural pressure to be thin: There are even alarming reports of parents who restrict their infant's food, hoping to prevent obesity later in life.
The connection between the weight-loss problems of people like Cathy and their history of repeated dieting cried out to be tested in a systematic way. But it's unethical to ask people to lose and regain a lot of weight repeatedly; it could be harmful. In addition, it would require many years to follow a group of human dieters long enough to see the natural consequences of yo-yo dieting. Our initial choice, then, was to look to the animal laboratory.
I enlisted the aid of Eliot Stellar, a physiological psychologist at the University of Pennsylvania; Marcie Greenwood, a biologist at Vassar College who ran one of the world's most productive labs for the study of obesity; and Eileen Shrager, a postdoctoral fellow in our lab.
We studied several groups of adult male rats, putting one group through an experience similar to human yo-yo dieting. These rats were put on a high-fat diet and became obese. Then they were placed on a balanced weight-loss diet until they returned to normal weight. After that, they were given free access to the high-fat food, and they regained their lost weight. This cycle was repeated so that our yo-yo dieters completed two cycles of loss and regain.
46 Days to Lose Excess
The results were striking. The yo-yo dieters needed 21 days to lose their excess weight during the first cycle. But after they regained this weight and started the second cycle, it took them 46 days to lose it again, even though their diet was exactly the same.
There was an even greater difference in the time it took the rats to put the weight back on. In the first cycle, the animals needed 45 days to return to their obese weight after they came off their diet. To regain the same weight in the second cycle took just 14 days. Weight loss was two times slower and regain was three times faster during the second round of yo-yo dieting than during the first round.
It appeared that our animals were responding to dieting by using food more efficiently. They gained more body weight per gram of food eaten, and so maintained their weight on fewer calories. This formed the cornerstone of our "cycling hypothesis." We thought that weight loss and regain enhanced the efficiency of food use and that weight would be lost more slowly and regained more rapidly with successive cycles.
Since the body needed fewer calories, the same number of calories that led to stable weight, or even weight loss, before dieting could produce a weight increase afterward. Weight cycling could actually contribute to obesity.