For people who are fit and thin when they were young, I think its natual for them to gain some weight as they age due to slower metabolism. I have a friend who used to be 5'3"-110lb, and is now 120 lb with no changes in the way she eats. As for me, I think I am thin except for the tommy area. I inherited my dad's stomach, so all the fat seems to concentrate in one area. As for people who are obese since childhood, I found a good section from a Biochemistry Book (Voet & Voet 2nd Ed).
Most obese people find it inordinately difficult to lose weight, or having done so, to keep it off. Yet most animals, including humans, tend to have stable weights; that is, if they are given free access to food, they eat just enough to maintain this so-called "set point" weight. The nature of the regulatory machinery that controls the set point, which in obese individuals seems to be aberrantly high, is just beginning to come to light.
Formerly grossly obese individuals who have lost at least 100 kg to reach their normal weights exhibit some of the metabolic symptoms of starvation: they are obsessed with food, have low heart rates, are cold intolerant, and require 25% caloric intake than normal individuals of similar heights and weights. In both normal and obese individuals, some 50% of the fatty acids liberated by the hydrolysis of triglycerols are reesterified before they can leave the adipocytes. In formerly obese subjects, this reessterification rate is only 35-40%, a level similar to that observed in normal individuals after a several day fast. The fat cells in normal and obese individuals, moreover, are of roughly the same size; obese people just have more of them. In fact, adipocyte precursor cells from massively obese individuals proliferate excessively in tissue culture compared to those from normal or even moderately obese subjects (adipocytes themselves do not replicate). Since fat cells, once gained, are never lost, this suggests that adipocytes, although highly elastic in size, tend to maintain a certain fixed volume and in doing so influence the metabolism and thus the appetite (I guess that would explain the problem areas). This insight, unfortunately, has not yet led to a method for lowering the set points of individuals with a tendency toward obesity.
Interesting research from UM-Ann Arbor.
http://www.med.umich.edu/opm/newspage/2004/wnt10b.htm
Mice with hyperactive Wnt10b gene eat all they want, but have half the body fat of normal mice