I recently read an article by Duffey and Popkin (2013) published in the American Journal of Preventive Medicine. The researchers attempted to quantify the effects of three suspected causes of the rise of obesity in children and adolescents. They did a similar study in 2007 on adults. The results are not definitive, but the three factors are worth noting.
Duffey and Popkin looked at changes in 1) the number of eating occasions (meals+snacks) per day, 2) the energy density of foods and beverages in those meals, and 3) the portion size of each item. My immediate reaction to their expectation that increases in portion size and number of times people eat would explain the rise in obesity was, "not necessarily."
A person might not gain weight under those conditions if they applied the concept of Volumetrics (B. Rolls) or chose foods low in energy density as the Dietary Guidelines advise. In the thirty years that were studied, portion sizes and number of eating occasions did rise (and sometimes fall). The energy density of the meals did not show a clear pattern of increase or decrease.
In the past 13 years, my number of eating occasions and portion size also increased, but my weight did not. That is because I have strictly adopted the low energy density concept. I was able to increase the size of my meals because I reduced the calories per gram in each of them. (I also eat no less than 8 times a day)
What the researchers did find when studying adults was that in the past 30 years, the change in total energy intake that is associated with the rise in obesity, has involved all three. More of the change is explained by portion size and an increase in number of meals eaten. In the last five or more years, portion sizes have begun to decrease.
You can review the abstract (a summary of what they did and what they found) by clicking here.