It doesn't say normal weight. It says previous set point. As you gain
weight you ratchet up your set point. You body defends the higher weight.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstra
ct&list_uids=15005834
http://ajpregu.physiology.org/cgi/content/full/287/6/R1297
http://www.findarticles.com/p/articles/mi_m1175/is_n6_v23/ai_7722351
http://www.foodproductdesign.com/archive/1995/0295CS.html

http://arbl.cvmbs.colostate.edu/hbooks/pathphys/endocrine/bodyweight/leptin.html

Basically, leptin is secreted by adipose tissue (fat). When leptin
levels drop below a certain level, you get hungry. Drop to far, you get
really hungry. Here is the killer: leptin falls, when your fat cells
begin to lose too much of their fat stores, so losing weight (adipose
tissue specifically) causes leptin to drop, and you get hungry all the
time--I think of this as that weird body hunger I get when my weight get
down to where I want it to be.

I think that it would be difficult to construct any experiment that
would prove existence of set points. You are quite right to be
skeptical. It is a fuzzy concept, unfortunately, but it does not
necessarily make it automatically worthless.

The best experiment about set points that I know of, is the Vermont
prisoner feeding experiment. I could say that they had very difficult
times raising their weight above their setpoint.

======================================================================

`` Sims was interested in whether the metabolic differences observed
between fat and thin people were the result or the cause of their body
type. Put simply, he wanted to know whether people are born fat or
made fat. He decided that the best way to sort this out was to
convince a group of slim volunteers to eat themselves fat and to
observe what happened to them when they reduced to their original
weight.

Sims was fortunate to have nearby a ready source of experimental
subjects: the inmates at Vermont state prison, sufficient numbers of
whom were willing to gorge themselves for science. At first the
prisoners proved enthusiastic trenchermen, as much as doubling their
usual daily intake of food. But as they fattened, they became
increasingly reluctant to overeat. Most found it extremely difficult
to gain weight, and eventually some started to drop out of the
study. Only 20 made it through the requisite 200 days, achieving an
average weight gain of 20-25lbs. Relieved of the high-calorie,
low-exercise regimen, all but two of the inmates quickly dropped the
newly acquired ballast. The pair of inmates who found it most
difficult to lose weight were those who had experienced the least
difficulty gaining weight in the first place. It was later discovered
that both these men had a family history of obesity.

equipped to balance energy intake and output, and to reach an energy
equilibrium, or "homeostasis", at which it felt naturally
comfortable. What was particularly interesting was that body weight
seemed somehow fixed, and was in most subjects resistant to change
over the short term. The prisoners with obesity in their backgrounds
were, it seemed, genetically inclined to reach homeostasis at a higher
weight than were others; the high-calorie diet only helped manifest
their genetic proclivity.''