The Potato Chip Study
Jacob Schor, ND, FABNO
June 28, 2011
Another fascinating paper by Dariush Mozaffarian and his team of researchers was published last week in the June 23rd issue of the New England Journal of Medicine. Typically we were to researchers by last name only but in the case of this particular researcher, I find pleasure in saying the full name. You may recall my recent mention of Dariush Mozaffarian’s studies on fish and heart disease. He and his colleagues have developed a certain expertise on sifting through large volumes of data and finding small but significant associations between dietary habits and the health changes associated with them.
This most recent study looked at the relationship between changes in lifestyle factors, particularly foods consumed routinely, and weight change. Participants were evaluated at 4-year intervals, with multivariable adjustments made for age, baseline body mass index for each period, and all lifestyle factors simultaneously.
The researchers followed 120,877 people including 22,557 men, who were free of chronic diseases and not obese at baseline in this study. The participants were enrolled in one of three separate cohort studies, either the Nurses’ Health Study (NHS), the Nurses’ Health Study II or the Health Professionals Follow-up Study. The follow-up periods were from 1986 to 2006, 1991 to 2003, and 1986 to 2006. This yielded 1,570,808 person-years of follow-up data to analyze.
The dietary factors assessed included fruits, vegetables, whole grains, refined grains, potatoes (including boiled or mashed potatoes and French fries), potato chips, whole-fat dairy products, low-fat dairy products, sugar-sweetened beverages, sweets and desserts, processed meats, unprocessed red meats, fried foods, and trans fats. Also evaluated were nuts, 100%-fruit juices, diet sodas, and subtypes of dairy products and potatoes. Time spent watching television, types of alcohol drinks consumed and time spent exercising were also followed. Positive and inverse associations were calculated for each cohort and expressed as age related change or multivariable associated change. The effect on weight of consuming specific foods was more pronounced when the only variable analyzed was age. Analyzing multiple variables blunted the effects yet they were still significant.
The average weight gain was 3.35 lb during each 4-year follow-up period or a weight gain of 16.8 lb over a period of 20 years.
Increased or decreased consumption of specific foods was associated with either weight gain or weight loss over each 4-year period. These effects can be expressed both as age adjusted change or multivariable adjusted change both expressed in pounds. The later calculations are usually less dramatic as the multiple variables tend to blunt the calculated effect. The study abstract and discussion use the values obtained calculated for multi-variable adjusted changes. For the point of discussion with patients, using the age adjusted change results may make a greater impression. Below, is a simple chart of the data. I’m not sure how well this will come through via our email mailing list so let me summarize how to interpret these numbers. Let’s look at fruit. If a person increases their fruit consumption by one serving a day, this dietary change is associated with a .69 pound decrease in their weight over a 4-year period. That’s the aged adjusted change. If we look at the right hand column for multivariable adjusted change, predicted weight loss would be less, a mere 0.49 pound weight loss.
Move down the chart and you will see why this study is jokingly referred to as the potato chip study. Eat a bag of potato chips a day and in a couple of years you’ll be close to a pound and three quarters heavier. Eat a serving of French fries and you’ll be more than 3 pounds heavier.
Inverse associations, that is weight loss per serving consumed, were seen for a number of foods. The most dramatic effects were seen in those who added yogurt and nuts into their diets. Daily consumption of nuts, was associated with slightly more than a pound of weight loss (-0.57 lb) over a 4-year period, and yogurt, a -0.82 lb loss. These were all linear relationships for both increased or decreased consumption. That is the weight change seen for increased consumption reversed along the same line if consumption decreased. For example, eating more French fries was associated with increased weight while eating less fries was associated with weight loss.
A serving a day of low fat yogurt will knock eight tenths of pound off you. A serving of fruit will knock off another half a pound and a serving of nuts another generous half pound. Thus according to Dariush Mozaffarian, eating a bowl of yogurt with fruit and nuts will slowly but surely knock two pounds off your body weight.
Food: increased dietary intake Weight Change within each 4 year period
Age Adjusted Change Multivariable adjusted change
Fruit -0.69 -0.49
Vegetables -0.25 -0.22
Nuts -0.78 -0.57
Whole fat dairy +0.25 +0.10
Butter +0.47 +0.30
Cheese +0.13 +0.02
Low fat yogurt -1.16 -0.82
Potato chips +3.01 +1.69
Potatoes +2.14 +1.28
French fried potatoes +6.59 +3.35
Whole grains -0.59 -0.37
Refined grains +0.56 +0. 39
Sugar sweetened beverages +1.32 +1.00
Sweets or desserts +0.65 +0.41
Processed meat +1.76 +0.93
Trans fats +1.44 +0.65
The length of time participants slept on average each night yielded a U-shaped association with weight gain; people who slept less than 6 hours or more than 8 hours gained more weight. An additional hour watching television led to a weight increase of 0.31 lb. Those participants who quit smoking experienced a weight gain of 5.17 lb during the first 4-year period in which they quit, but after that initial gain, the former smokers gained only 0.14 lb in the following 4-year period.
Implications: This study confirms what we have believed for years and changes little, if anything, in our practices. We already teach our patients to eat more vegetables, fruits, nuts, whole grains and yogurt. We tell them to avoid simple carbohydrates, sugars, refined grains and desserts. We especially tell them to avoid regular consumption of fried simple carbohydrates like French fries and potato chips. Certainly having these numbers to present to our patients with a Harvard address and New England Journal of Medicine logo attached lend strength to our message but they do not change it
These data squash the notion that weight gain or loss is a simple equation of caloric intake versus energy expenditure. Weight gain or loss is more complex and increasing consumption of ‘healthy foods’ and avoidance of bad foods, especially it would seem, potatoes, leads to weight loss. This concept is not new to any of us. Simple carbohydrates increase weight gain.
Perhaps the most important thing about this study is that the food and weight change lists and especially the graphic charts that sum up these data may become effective tools for educating patients.
Reference: Mozaffarian D. Tao Hao PH. Rimm EB.Willett WC. Hu PH. Hu FB. Changes in Diet and Lifestyle and Long-Term Weight Gain in Women and Men. N Engl J Med. 2011 Jun 23;364(25):2392-404.
A pdf of the study is posted at: