Readers warning: what follows is a brief review of some of the flaws in nutrition research. If you're like me, your eyes glaze over whenever you try to get through study results, and frequently you end up with a headache and no more enlightened. Please try to bear with me because I hope I put this information in plain English, and that you actually will be enlightened by this discussion.
A new study published in the April issue of the
Archives of Internal Medicine suggests that the "DASH" diet
reduces women's risk of heart disease and stroke, and lowers their blood pressure. "DASH " stands for Dietary Approaches to Stop Hypertension. The DASH diet is comprised of lots of fruits, vegetables, and plant proteins (beans and nuts), and is low in animal protein. It includes a moderate amount of low-fat dairy products.
The DASH diet has previously been shown to lower blood pressure and reduce "bad" LDL cholesterol. Both are factors in the risk of developing cardiovascular disease (CVD) and stroke. Teresa T. Fung, ScD, of Simmons College, Boston and colleagues studied the eating patterns of 88,517 female nurses aged 34 to 59 to determine if sticking to a DASH diet affected a woman's risk of such diseases. The women did not have CVD or diabetes when the trial started. This type of study is called an "observational" study; it takes a group of individuals and follows some aspect of their life and how it influences their health over time.
In their study the researchers followed the nurses for 24 years. Over that time they asked the women to report what types of foods they regularly consumed over the previous year. The researcher's then divided the foods into specific categories, and assigned those categories a DASH score. The more fruits, vegetables, whole grains, nuts, and beans the woman ate and the more she adhered to the recommended low-fat dairy choices, the higher her DASH score became. Scores were lower in women who consumed more red and processed meats, salt, and sweetened drinks. Women with the highest DASH scores had a 24% reduction in heart disease and an 18% lower risk of having a stroke when compared to the women with the lowest DASH scores. The authors point out that the women with the highest DASH scores also appeared to live overall healthier lifestyles. They were less likely to be current smokers, more likely to exercise, and tended to consume high amounts of fiber and omega-3 fatty acids than the other study participants. This is a problem with an observational group; the study design doesn't control for other things that might influence the study's outcomes.
I mentioned here recently that I am currently reading
In Defense of Food, by Michael Pollan. In this book Mr. Pollan points out, with numerous examples, how nutrition science is very flawed. One of his examples is the Nurse's Health Study (Belanger, C.F., Hennekens, C.H, Rosner, B., et al. "The Nurses' Health Study."
American Journal of Nursing. (1978):1039-40), which is not the same study currently being discussed by Dr. Fung. The Nurses' Health Study of 1978 was also a long-term observational study that collected data on eating habits and health outcomes of more that 100,000 women over several decades. It is still considered one of the best studies of its kind, yet it has significant flaws.
One of these is it's reliance on food frequency questionnaires, which appear to have been used in this DASH diet study as well. A food frequency questionnaire asks an individual to recall what they have eaten over a certain period of time. In the 1978 Nurses' Health Study the questionnaire asked such things as "Did you eat chicken or turkey in the last 3 months?" and "When you ate chicken or turkey, how often did you eat the skin?" Other questions asked about vegetable intake, and whether or not the vegetable was steamed or fried, and if fried, in what kind of fat.
Now think about your diet for a minute. When was the last time you ate french fries? When you ate them, did you fry them yourself, or did you eat them at a restaurant? If cooked at a restaurant, what are the odds that you know what kind of fat they used to cook your fries? What are the odds that you recall how many times you ate french fries in the last 3 months? How about the last year, which was the period of recall for the DASH study? Do you see how unreliable this data might be?
There are several problems with relying on individuals to report their food intake, including the difficulty with recall and knowing how things were cooked, as I pointed out above. Another problem is that people don't tell the whole truth. Scientists know that people underestimate their food intake all of the time; they have even developed scientific figures to measure the degree of error that occurs with self-reporting of dietary intake. The studies that assess the validity of the food frequency questionnaires indicate that on average people eat 1/5 and 1/3 more volume than they indicate on these questionnaires, and it is likely that this underestimates the problem. This is because scientists know that there are 3,900 calories of food produced each day in America for each person living here, yet we admit to only eating 2,000 per person. So although there is some food waste, it can't be enough to make up the other 1900 calories per person and someone
is eating them!
To try to improve the odds that a person's recall is accurate, scientists match the results of the food frequency questionnaires with the results of an individual's recall of their intake in the last 24 hours, which is thought to be somewhat more reliable. But again, think about what you ate in the last 24 hours. How typical of your "usual" diet was this intake? Personally, I ate fried onion rings tonight, from a restaurant. I rarely eat fried foods, and I have no idea what kind of oil they used. So this is not in any way typical of my intake, and would totally skew my data if this were the day of my 24 hour recall.
Research into how we eat and how our food intake influences our health are full of potential errors; the examples above are a portion of those errors. Let me tell you about one more study to really make this point clear.
My mom participated in the "
Women's Health Initiative" study, specifically in the
Dietary Modification arm of it. She was randomly assigned to the group that was supposed to decrease their fat intake to 20% of their total daily calories, increase their consumption of fruits and veggies to 5 or more servings a day and whole grains to 6 or more servings per day. They monitored their food intake (through those dietary recall sheets), attended nutrition education sessions monthly for the first year and then four times per year for the remainder of the study. The comparison group kept us their usual eating habits, received only standard nutrition guidelines, and filled out health records every six months. This was a large group, 48,800 women, who were followed for 8-12 years.
The researchers believed that they had removed the bias that was a problem with the DASH study because they randomly assigned women to the two groups, modified diet versus usual diet. This would divide the women who tended to eat better, exercise more, and not smoke into both groups and this would result in those factors having equal influence on the outcome of both groups. This was supposed to mean that the only factor being measured was the change in diet versus the "usual" diet of the control group. But once again nutrition "science" fails us. In addition to the problems outlined above about basing data on dietary recall, consider this: women in the dietary modification arm of the study got their dietary fat intake down to about 24% at their best point in the study period and at the end their fat intake had drifted back closer to 29%.
During that same time, women in the control group were hearing in the media that "low fat" was the way to go (WHI enrollment began in 1991 and the entire study period from enrollment to completion lasted 15 years; during that time the "low-fat diet" came into mainstream media discussions and into eating vogue). Hearing this low-fat mantra meant that the "controls" began to decrease their fat intake resulting in a reduction in fat in their diets that approached the percentage of dietary fat in the study group. This influence is called the "treatment effect." Additionally there was no effort made to control where the fat came from, so women who get their dietary fat from animal sources were lumped into the same group as those who get their fats from things like soy products, avocados and olive oil! When the results were announced in 2006 the New York Times printed the headline, "
Low-Fat Diet Does Not Cut Health Risks, Study Finds." This is because in the end there was likely very little difference between the diets of the two groups!
I hope that this little research lesson has both enlightened you on a complicated subject and led you to a few of the same conclusions that I have come to: nutrition "science" is fraught with research errors and it's no wonder we're confused about what to eat!
Frankly I think I'm going to follow some of Michael Pollan's simple advice, the first of which is stated on the front of
In Defense of Food:
"Eat Food. Not Too Much. Mostly Plants."
How does he define food? For that you'll have to read this really great book.
Bon Appetite! Laurie
Related Topics: Technorati Tags: DASH diet, heart health, Michael Pollan, In Defense of Food, nutrition research
