Changes in diet and lifestyle that decrease weight gain

This week's New England Journal of Medicine has an article that has been widely reported on the TV news and that sounds suspiciously exactly like what your mother has been telling you. (1)
Investigators at Harvard took data from 120,877 nurses and male health professionals who were not sick or obese to begin with and who have been followed in The Nurses' Health Study (NHS) and The Nurses' Health Study II (NHSII) and The Health Professionals Follow-up Study (HPFS) over many years from 1986 to 2006 so far. Every 4 years these mostly educated whites were asked lots of questions about what they ate and other life style habits and this information was correlated with how much weight they gained. So they were able to analyse what happens when these diet and lifestyle habits changed over the course of some 20 years.
Within each 4-year period, participants gained an average of 3.35 lb for a total of 16.8 lb over the 20 years. The younger nurses enrolled later gained much more than the average. So what should we eat? Well the differences were small but highly statistically significant since they were working with so many data points involving over a million and a half person-years of follow-up. On the basis of increased daily servings of the dietary components checked, some of the bigger 4-year weight increases were associated with: potato chips (1.69 lb), French fries (3.35 lb), baked potatoes (0.57 lb), and sugar-sweetened beverages (1 lb). Less weight gain was associated with intake of: vegetables (-0.22 lb), whole grains (-0.37 lb), fruits (-0.49 lb) nuts (-0.57 lb) and yogurt (-0.82 lb). Other lifestyle factors that were independently associated with statistically significant weight change were physical activity (-1.76 lb) and TV watching. (0.31 lb per hour per day).
So in addition to "eat less and exercise more" which is notoriously unhelpfull, can we say that you should eat certain things and not watch TV? Well that's what this report would seem to indicate but it's tricky. Total Calorie intake is not well estimated from dietary questionnaires like used in this study. It may be that people are more likely to be dieting if they are doing all the other things that their mother told them to do. It could be that eating more or less of any kind of food or beverage may change the total number of Calories that you eat. That is, that dietary quality influences dietary quantity. But it also suspicious for being reverse causality. People who eat "quality" food are dieting.
Then there are other anomalies in the findings. Why would french fries be twice as bad as potato chips and many times worse than baked potatoes. They are all the same molecules. Potato chips have 150 Cal/oz and 10 grams fat, French fries have 95 Cal/oz and 5 grams fat, and a baked potato with the skin and 2 teaspoons butter has 38 Cal/oz and 8 grams fat. I guess when you eat potato chips you're usually drinking beer and watching a football game on TV but when you eat a baked potato you are at dinner in a fine restaurant. I don't know why french fries are the worse than potato chips. Maybe they taste so good that you just eat much more. Why is eating yogurt so good but drinking skim milk or whole milk makes no difference. Sugar sweetened beverages don't have any more sugar than 100% fruit juice but had a 3 times bigger weight gain.
The authors of this report have a long discussion of the implications and anomalies of this data but it doesn't really hang together very well. The only way to be sure is take a bunch of people and make them eat "right" and another bunch and make them eat "wrong" and see what happens. But that's really hard.
At least the dietary recommendations that would come from these findings, if you add salt restriction, which not eating potato chips and french fries would be part of, then you have the DASH diet. Remember the DASH diet recommendations? -Focus on fruits, vegetables, and fat-free or low-fat milk and milk products. Emphasize whole grains, fish, poultry, beans, seeds, and nuts. Limit sweets, added sugars and sugary beverages, and red meats. Though the evidence for weight reduction or control is a little circumstantial, the evidence that it will lower your blood pressure is very good.

1. Mozaffarian D et al. Changes in diet and lifestyle and long-term weight gain in women and men. N Engl J Med 2011;364:2392-404.