Dietary Patterns1

Dietary Patterns1

Katherine L. Tucker

Although much of the history of nutrition has involved the study of individual nutrients, implications of deficiency, and determination of their mechanistic roles and requirements, it is foods that provide our nutrients and whole diets containing the wide array of nutrients and phytochemicals that work together to influence our health. We have learned much about the important actions of individual nutrients from in vitro studies, animal studies, and human metabolic studies and have conquered many deficiency diseases with fortification or supplementation. Based on these often short-term results, we have also estimated optimal nutrient exposures for prevention of specific outcomes, including heart disease and cancers. These assumptions have then been tested in epidemiologic studies, which have sometimes supported the original findings, but sometimes have not.

The next step in this scientific process has been to move to large-scale clinical trials for those with highly consistent results, usually by providing nutrient supplements to test their efficacy in preventing disease. Notable examples of these studies include several high-profile trials of β-carotene trials in the United States. Based on highly consistent results from many diverse studies, β-carotene was expected to contribute to prevention of lung cancer. In complete contrast to expectation, however, these trials found that the supplement tended to have no effect relative to placebo or, in some cases, actually to have an adverse effect. In the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) trial, with 29,133 male smokers, β-carotene supplements were, rather than protective, associated with an 18% increased incidence of lung cancers and 8% increased overall mortality (1). Similarly, the Beta-Carotene and Retinol Efficacy Trial (CARET), in 18,314 middle-aged highrisk men and women, showed a 28% higher lung cancer incidence and 17% higher mortality (2), although a third β-carotene trial, within the Physicians’ Health Study, found no effect on mortality in younger men, at lower risk for lung cancer (3). This medical model approach to nutrition has led to several surprises, which have opened our thinking to the need to move nutrition research back to diet and foods and beyond that, to complete dietary patterns for improving overall health.

Even when moving to dietary interventions, results have often been disappointing when they are focused primarily on targeted specific nutrients. For example, much of the nutrition advice in the late twentieth century focused on low-fat and low-cholesterol diets for prevention of cardiovascular disease (CVD), based on many consistent short-term studies of intermediate metabolic markers. This focus led to several behaviors that, in retrospect, were not necessarily optimal for overall health, including the development of low-fat products that increased exposure to refined carbohydrates and excessive avoidance of otherwise healthy foods, such as eggs.

One of the largest dietary intervention trials to be conducted, the Women’s Health Initiative (WHI), with 48,835 postmenopausal women followed up for 8.1 years, focused primarily on changing to a low-fat diet—to 20% of calories—albeit with increasing fruit and vegetable intakes to 5 servings per day and “grains” to 6 servings per day (4). The intervention led to reported decreases in fat intake by 8.2% of total energy, increases of fruit and vegetables by 1.1 servings, and increases in grains by 0.5 servings per day.
However, the diet had no significant effect on coronary heart disease (CHD), CVD, or stroke incidence. Since then, there has been an increasing realization that the focus on low total fat intake was misguided and that the effect of differing fatty acids on health is complex and involves a balance of healthy fats rather than avoidance of all fats.

Further support for examination of the total diet includes methodologic limitations of the single food or nutrient approach (5). These limitations include collinearity among nutrients that makes it difficult to be sure whether one nutrient is having an effect or whether it is confounded by another in the same diet. This may lead to either overestimation or underestimation of an association, because adjusting for another collinear nutrient may mask a true association. A well-known and simple example of this is shown by the conflicting hypotheses that high dietary fat was associated with colon cancer and that, conversely, dietary fiber may be protective. Because most diets high in fat are low in fiber, early results were nonconclusive (6). More important, the known interactive effects of many nutrients are likely to add to more than the sum of their parts, and these effects can be fully evaluated only when considering the full diet. As understanding of nutrient mechanisms have become more sophisticated, there is increasing recognition that, beyond correcting deficiency, large doses of individual nutrients taken out of their natural food matrix do not often perform the same way as when these nutrients are obtained in natural form within the food matrix, with other nutrients and many phytonutrients that interact with them.

Together, this evolution of diet and health research has led to an intensified examination of total dietary patterns to gain insights into how what we eat affects health outcomes (7). By their nature, dietary patterns are complex and may be measured in many different ways. Approaches to measurement have evolved over time from simple adequacy scores and indices to evidence-based scores and, most recently, to empiric patterns existing in the population (Table 105.1).


Early dietary pattern scores were straightforward and nutrient based. These include the nutrient adequacy ratio (NAR), which is defined as the average intake per day of a nutrient divided by the recommended intake, and the mean adequacy ratio (MAR), defined as the sum of the NARs, divided by the number of nutrients included (8).

Because different foods contain different nutrients, it has long been a foundation of nutrition recommendations to include a variety of foods. Therefore, another score that has been used effectively for some time is that of dietary diversity or variety. Consuming a variety of foods has been linked clearly, in many contexts, with dietary quality (9). The concept of dietary diversity has long been used in developing countries, where it has been a reliable indicator of nutritional status in situations where food is limited and based heavily on starchy staples (10). This concept has been revived more recently in the study of diets in developing countries, where dietary diversity has consistently been shown to relate to child growth (11, 12).




Nutrient adequacy ratio

Adherence scores from


Principal Components

Dietary diversity or variety



Groups of individuals from

Healthy Eating Index

cluster analysis

Alternative Healthy Eating


Dietary Approaches to

Stop Hypertension

(DASH) diet

Mediterranean diet score

Dietary diversity, or variety, has been measured in different ways, including a simple count of differing foods or as the number of food groups consumed over a specified period of time (13). In the United States, variety has been shown to be associated with diet quality (14) as well as with risk of chronic disease and total mortality (15, 16). In contrast, in situations where food choices are plentiful, investigators noted that a simple count of the variety of foods may be related to high energy intake and obesity. A closer look shows that this association resulted mainly from variety in sweets, snacks, and carbohydrates, whereas variety in vegetables was associated with lower body mass index (BMI) (17). However, even in the context of developed countries, variety assessed by total number of different food items has been associated with improved nutrient intake profile in older adults (18).

An analysis of national survey data showed that—whether defined as a count of basic commodities consumed, a count of food codes reported, a count of 5 Food Guide Pyramid (FGP) food groups consumed, or a count of 22 FGP subgroups consumed—all types of dietary variety scores were associated positively with mean nutrient adequacy across 15 nutrients, and they were associated inversely with intake of added sugars and saturated fat. The strongest associations were seen with the measure using 22 food groups (19). Use of food group measures in this way supports the concept of consuming a true diversity of foods. Other current measures focus on recommended foods only. Variety of fruit and vegetables has, for example, been associated with lower markers of inflammation (20).

More complex dietary pattern scores are based on multiple dietary guidelines. In the United States, the US Department of Agriculture (USDA) Healthy Eating Index (HEI) (21) is a score based on degree of adherence to the FGP and US Dietary Guidelines (USDG). It is revised by the USDA periodically, with updated information (22) (,
and it has been reinterpreted by others to include new scientific information (23). The original score had 10 items, each with a subscore of 1 to 10, to measure the degree to which each criterion was met (servings of grains, vegetables, meat, and milk, as well as recommended intakes of total fat, saturated fat, cholesterol, and sodium, and variety).

Among other things, the HEI has been associated with lower obesity (24), lower abdominal obesity (25), lower depression score (26), better physical performance (27), reduced age-related nuclear lens opacities (28), and risk of heart disease (29). In addition to the original components, the Alternative HEI (AHEI) emphasizes choosing unsaturated fat over saturated or trans-fats and inclusion of cereal fiber, fish, and moderate alcohol consumption in the diet. In several studies, the AHEI was shown to be more strongly related to several health outcomes than was the original HEI (see the chapter on foundations of a healthy diet), and it has been associated with lower risk of type 2 diabetes (30), CVD (31), colorectal cancer (32), and, most recently, with mortality in Britain (33).

Condition-Specific Designed Dietary Patterns: The DASH Diet

One of the most successful dietary interventions to be introduced is the Dietary Approaches to Stop Hypertension (DASH) diet. This approach used available evidence from earlier research to design a diet that could help to reduce blood pressure (34). The diet emphasizes fruit, vegetables, whole grains, low-fat dairy products, fish, chicken, and lean meats and was designed to be low in saturated fat and cholesterol, moderately high in protein, and high in potassium, magnesium, calcium, and in fiber.

The initial DASH study compared this diet with one that resembled the current US typical pattern and with one that emphasized only fruit and vegetables. In contrast to the WHI, which gave dietary advice to meet a prescribed low-fat pattern, the DASH study initially provided food to participants to ensure compliance. After 8 weeks, reductions in systolic and diastolic blood pressure were significantly greater among those following the DASH diet than among those on the control diet by 5.5 and 3.0 mm Hg, respectively. The DASH diet also showed greater reductions in blood pressure than did the diet emphasizing fruit and vegetables alone by 2.7 and 1.9 mm Hg (35). Results with other indicators showed that the diet also led to improvements in plasma lipids (36), homocysteine concentration (37), and Framingham risk score (38).

After the original trial, the DASH diet was tested with the addition of sodium reduction (DASH-Sodium), in which the combination of both was shown to reduce blood pressure more than either approach alone (39). A secondary analysis also showed that beneficial effects of the DASH diet led to reductions in the bone turnover markers serum osteocalcin (by 8% to 11%) and C-terminal telopeptide of type I collagen (by 16% to 18%), with further reduction in osteocalcin and urinary calcium excretion with the addition of low sodium (40). A 1-year follow-up showed sustained benefits in dietary behavior and blood pressure (41). A replication of the DASH diet in Iran resulted in improvements in blood pressure and also in body weight, fasting blood glucose, glycosylated hemoglobin, and low-density lipoprotein (LDL) cholesterol. Increases in high-density lipoprotein (HDL) cholesterol were also seen in adults with type 2 diabetes who were following the DASH diet (42).

The DASH diet uses a total diet approach. Its efficacy was clearly shown when food was provided, and, since then, several studies have launched community-based interventions with diet education and encouragement to demonstrate effectiveness in a larger context. The ENCORE study tested the diet with overweight or obese adults with higher than normal blood pressure in a hybrid form: the investigators provided food for 2 weeks, followed by weekly education and motivational sessions to continue the diets independently. Participants were randomized to the DASH diet, the DASH diet plus exercise and weight management, or usual care. After 4 months, the DASH-only group had significant improvements in blood pressure (declines of 11.2 and 7.5 mm Hg [systolic and diastolic] versus 3.4 and 3.8 mm Hg in the control group). The DASH plus exercise group had even greater declines of 16.1 and 9.9 mm Hg, respectively (43).

The PREMIER trial was designed to test the diet with an advice-only intervention. Older adults with higher than optimal blood pressure were randomized to one of three groups: a standard advice-only control, an established behavioral intervention for weight loss and sodium reduction, or the same behavioral intervention plus the DASH diet menus and instruction. After 6 months, there was a graded response, with change from baseline prevalence of hypertension of 38% to 26%, 17%, and 12%, respectively, in groups 1 to 3. The largest effect was seen with the DASH diet, although it did not differ significantly from the basic behavioral intervention (44). The DASH group did report greater increases in intake of fruit, vegetables, and low-fat dairy products than other groups, but apparently the changes were not sufficient to see a statistically significant additive effect on hypertension over the weight loss and sodium reduction in the basic intervention. A subsequent analysis, however, showed that only the DASH group showed significant improvements in insulin sensitivity (45).

Given the success of the controlled studies, data from several cohort studies have been examined for adherence to the DASH diet by creating scores from reported free-living diets. In Sweden, women in the top quartile of the DASH diet score had 37% lower 7-year incidence of heart failure than did those in the lowest quartile (46). In a parallel study, Swedish men in the highest quartile of the DASH score had 22% lower incidence of heart failure than did those in the lowest quartile (47). Adherence to the DASH diet has been associated with lower risk of
heart disease and stroke in US women (48), and with lower incidence of type 2 diabetes in US adults. In the Insulin Resistance Atherosclerosis Study (IRAS), adults in the highest, versus the lowest, tertile of DASH adherence score were 69% less likely to develop incident diabetes over 5 years (49). However, one cohort study using a DASH score did not find associations with hypertension or cardiovascular mortality in a large cohort of US women; the investigators noted that adherence to the score may have been too low to see an effect (50).

DASH diet scores have also been examined in youth. An intervention with adolescents with prehypertension or hypertension who were randomly assigned to standard of care or to a DASH diet intervention showed that those in the DASH group had greater declines in systolic blood pressure (51). Among youth 10 to 22 years of age with type 1 diabetes, those in the highest DASH diet score tertile had significantly lower blood pressure (52), lower LDL/HDL cholesterol ratios, and lower glycosylated hemoglobin than those in the lowest tertile (53). Prospectively, adolescent girls in the Prospective National Growth and Health Study had lower gains in BMI over 10 years, when in the highest, versus the lowest, quintile of DASH diet score (54).

Together, the evidence for the DASH diet in chronic disease prevention is strong. Although designed specifically for lowering blood pressure, this diet has been shown to have beneficial effects on other metabolic indicators, a finding suggesting that it is likely a beneficial behavior for health in most individuals. Given the general success of the DASH diet, it has been promoted for general population use by the National Heart, Lung, and Blood Institute (55), the Mayo Clinic (56), the American Heart Association (57), and in a popular book (58).

Dietary Patterns from Observed Healthy Populations: The Mediterranean Diet Score

The DASH diet is not the only approach to healthy eating, however. The AHEI, described earlier, was a modification of the original HEI to include additional evidence on diet and health, particularly aspects of an observed dietary pattern that has been associated with lower chronic disease risk relative to other regional patterns: the Mediterranean diet. This diet was first highlighted in 1970 by Ancel Keys in his seven-country study (59). CHD was noted to be lower in the Mediterranean region than other parts of the world, although total fat intake—in the form mainly of olive oil—was not low. In 1995, a Mediterranean diet pyramid was introduced, following an expert conference sponsored by Oldways Preservation & Exchange Trust (http:// and the World Health Organization/Food and Agriculture Organization Collaborating Center in nutritional epidemiology at the Harvard School of Public Health in Boston (60). This pyramid was based on the food patterns of Crete, other parts of Greece, and southern Italy in the 1960s (61). The 1960s Mediterranean diet was characterized by high intakes of fruit, vegetables, breads and cereals, potatoes, beans, and nuts and seeds, with olive oil as the principal source of fat; low to moderate amounts of cheese and yogurt, fish, poultry, and eggs; moderate amounts of wine, consumed mainly with meals; and low intakes of red meat. This diet is low in saturated fat (≤8% of energy), with total fat from 25% to more than 35%.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jul 27, 2016 | Posted by in PUBLIC HEALTH AND EPIDEMIOLOGY | Comments Off on Dietary Patterns1

Full access? Get Clinical Tree

Get Clinical Tree app for offline access