Part D. Chapter 2: Dietary Patterns, Foods and Nutrients, and Health Outcomes
A healthy diet is a pillar of well-being throughout the lifespan. It promotes the achievement of healthy pregnancy outcomes; supports normal growth, development and aging; helps maintain healthful body weight; reduces chronic disease risks; and promotes overall health and well-being. Previous Dietary Guidelines Advisory Committees focused on examining specific foods, nutrients, and dietary components and their relationships to health outcomes. In its review, however, the 2010 DGAC noted that it is often not possible to separate the effects of individual nutrients and foods, and that the totality of dietthe combinations and quantities in which foods and nutrients are consumedmay have synergistic and cumulative effects on health and disease.1 This approach has been adopted by others as well (e.g. American Heart Association, American College of Cardiology and the National Cancer Institute) and is being used by the 2015 DGAC. The 2010 Committee acknowledged the importance of dietary patterns and recommended additional research in this area. After the release of the 2010 Dietary Guidelines for Americans, the USDA Nutrition Evidence Library (NEL) completed a systematic review project examining the relationships between dietary patterns and several health outcomes, including cardiovascular disease (CVD), body weight and type 2 diabetes.2 Their report has been used by the 2015 DGAC.
As also noted in the 2010 Dietary Guidelines for Americans, individuals can achieve a healthy diet in multiple ways and preferably with a wide variety of foods and beverages. Optimal nutrition can be attained with many dietary patterns and a single dietary pattern approach or prescription is unnecessary. Indeed, for long-term maintenance, a dietary pattern to support optimal nutrition and health should be based on the biological and medical needs as well as preferences of the individual.
Dietary patterns are defined as the quantities, proportions, variety or combinations of different foods and beverages in diets, and the frequency with which they are habitually consumed. Americans consume many habitual dietary patterns, rather than a “typical American pattern,” which reflect their life experiences and wide-ranging personal, socio-cultural and other environmental influences. The nutritional quality of a dietary pattern can be determined by assessing the nutrient content of its constituent foods and beverages and comparing these characteristics to age- and sex-specific nutrient requirements and standards for nutrient adequacy, as shown in Part D. Chapter 1: Food and Nutrient Intakes, and Health: Current Status and Trends for the USDA Food Patterns, including the “Healthy U.S.-style Pattern,” the “Healthy Mediterranean-style Pattern,” and the “Healthy Vegetarian Pattern.” Understanding the array of dietary patterns in a population and their nutrient quality allows a more complete characterization of individual eating behaviors and enables their examination in relationship with diverse health outcomes. For these reasons, the DGAC focused on considering the evidence for overall dietary patterns in addition to key foods and nutrients. A major goal was to describe the common characteristics of a healthy diet, which informed and is complementary to the quantitative description of dietary patterns provided in Part D. Chapter 1: Food and Nutrient Intakes, and Health: Current Status and Trends.
Dietary patterns can be characterized in three main ways, drawing from Dr. Susan Krebs-Smiths presentation to the DGAC during the second public meeting (available at www.DietaryGuidelines.gov). The first is by the use of an a priori index that is based on a set of dietary recommendations for a healthy dietary pattern as a result of scientific consensus or proposed by investigators using an evidence-based approach. An individuals index/score is derived by comparing and quantifying their adherence to the criterion food and/or nutrient component of the index and then summed up over all components. A populations average mean and individual component scores can be similarly determined. Examples of dietary quality scores include: the Healthy Eating Index (HEI)-2005 and 2010,3 the Alternate HEI (AHEI) and updated AHEI-2010,4 the Recommended Food Score (RFS),5 the Dietary Approaches to Stop Hypertension (DASH) score,6 the Mediterranean Diet Score (MDS),7 and the Alternate Mediterranean Diet Score (aMed).8
The second method of dietary pattern assessment is through data-driven approaches, such as cluster analysis (which addresses the question, “Using the self-reported food and beverage intake data are there groups of people with distinct (non-overlapping) dietary patterns?”) and factor analysis (which addresses the question, “Which components of the diet track together to explain variations in food or beverage intake across diet patterns?”). These data-driven approaches are outcome-independent. That is, the relationships between the dietary patterns and intermediate or longer-term health outcomes are examined once the patterns themselves are defined. Other data-driven approaches are outcome-dependent, such as reduced rank regression (which addresses the question, “What combination of foods explains the most variation in one or more intermediate health markers?”).
The third method examines individuals food and beverage intake preferences as they are commonly defined by foods included or eliminated. In cohort studies, this pattern is usually based upon qualitative self-reported behaviors rather than detailed questionnaires. Vegetarianism and its various forms (e.g., ovo-lacto vegetarianism) are examples of this type of dietary pattern.
The dietary patterns approach has a number of major strengths. The method captures the relationship between the overall diet and its constituent foods, beverages and nutrients in relationship to outcomes of interest and quality, thereby overcoming the collinearity among single foods and nutrients. In so doing, it considers the inherent interactions between foods and nutrients in promoting health or increasing disease risk. Because foods are consumed in combinations, it is difficult, if not impossible, to determine their separate effects on health. Relationships or effects attributed to a particular food or nutrient may be accurate or reflect those of other dietary components acting in synergy. The dietary pattern approach has advanced nutrition research by capturing overall food consumption behaviors and its quality in relationship to health.
Despite these considerable strengths, however, the approach has several limitations that are important to consider. First, the dietary assessment instruments used to define the dietary patterns (e.g., food frequency questionnaires [FFQ] and 24hour or multi-day dietary recalls or records) are based upon self-report and may introduce levels of report bias that can attenuate diet-health relationships. The FFQ has been evaluated as a valid and reliable measure of usual food and nutrient intake. However, the extent to which data from FFQs are valid measures of dietary patterns is not well established. Second, dietary patterns are not uniformly defined by investigators and vary substantially from one study to the next even though studies may use the same nomenclature. This may hamper cross-study comparisons and limits reproducibility. Third, scoring algorithms used to evaluate dietary pattern adherence may differ and affect the results of studies examining specific health outcomes. Fourth, data-driven methods may not derive comparable patterns in different populations because these patterns may be population specific. Lastly, dietary patterns do not assess the frequency of meal and snack consumption, specific combinations of foods consumed together, and aspects of food purchase and preparation, all of which may influence the overall dietary pattern.
Another challenge to examining dietary patterns is that randomized dietary intervention studies have used different approaches for ensuring that subjects comply with the intervention diet when testing their relationships with health outcomes. For example, randomized controlled trials (RCTs), such as Prevencion con Dieta Mediterranean (PREDIMED), coached participants to follow a dietary pattern and provided them with key foods (e.g., olive oil or nuts) to facilitate adherence. In contrast, feeding studies (another form of intervention study), such as those conducted in the DASH and the Optimal Macronutrient Intake Trial for Heart Health (OmniHeart), provided all food to be consumed to each participant. These study designs across randomized trials and feeding studies provide strong evidence for the benefits and risks of particular dietary patterns because a prescribed intervention allows relatively precise definition of dietary exposures, and randomization helps ensure that any potential confounding variables are randomly distributed between study arms. However, some trials (i.e. DASH, OmniHeart) are necessarily restricted to testing a dietary patterns effect on an intermediate outcome or a surrogate endpoint, such as blood lipids, because of the complexities involved in maintaining dietary compliance over long study duration. Additionally, the feeding trials fail to represent what happens in real world situations. Thus, well-conducted observational cohort studies provide an important evidentiary complement to RCTs because they enable the study of hard endpoints for disease in addition to intermediate outcomes and often provide a wider range of exposures for study.
Dietary patterns and their food and nutrient characteristics are at the core of the conceptual model that has guided the DGACs work (see Part B. Chapter 2: 2015 DGAC Themes and Recommendations: Integrating the Evidence), and the relationship of dietary patterns to health outcomes is the centerpiece of this chapter. The Committee considered evidence about the relationship of diet with several health outcomes that are listed as major public health outcomes of concern in Part D. Chapter 1: Food and Nutrient Intakes, and Health: Current Status and Trends. Several of these outcomesCVD, overweight and obesity, type 2 diabetes, congenital anomalies, and bone healthalso were addressed by the 2010 DGAC. Otherscancers (lung, colon, prostate and breast) and neurological and psychological illnesswhile previously addressed, are considered here in more depth and represent an expanded list of health outcomes for which there is growing evidence of a diet-disease relationship. The 2015 Committee was not able to consider the relationship between dietary patterns during the peri- and prenatal period and pregnancy outcomes (e.g., birth weight, preterm birth, pregnancy complications) or other cancer outcomes, such as total cancer mortality or gynecological, pancreatic, and gastric-esophageal cancers due to time limitations and limited work done in these areas involving dietary patterns. However, it is important to note that recently the NIH-AARP Diet and Health Study (n = 492,823) conducted in the United States demonstrated that high adherence on several indices (the HEI-2010, the AHEI-2010, the aMED, and DASH) was associated with lower risk of overall CVD and cancer mortality.9 The authors concluded that this finding provides further credence for using the dietary pattern approach, indicating that multiple dietary indices reflecting core tenets of a healthy diet may lower the risk of mortality outcomes.9
Over the course of the DGACs review, when strong or moderate evidence related to dietary patterns and a particular health outcome was available, the Committee focused its discussion on dietary patterns and, as possible, highlighted the most consistent common food and nutrient characteristics identified in the dietary patterns literature. When only limited or insufficient evidence related to dietary patterns and a particular health outcome was available (as in the case of congenital anomalies and neurological and psychological illnesses), the Committee summarized these findings and also provided a brief summary of existing evidence on specific foods and/or nutrients and selected health outcomes.
In addition to its work on dietary patterns, the DGAC considered conducting an evidence review on the relationship between the role of the microbiome and various health outcomes. This novel area of research has generated considerable interest in the scientific community and the lay public. Investigators are examining the diversity of organisms (i.e., microbes) that inhabit different parts of the body such as the gut, mouth, skin, and vagina, and are attempting to understand how the microbial communities are influenced by diet, environment, host genetics and other microbes, as well as their association with various health outcomes. The DGAC conducted an exploratory search but did not find sufficient evidence to address this question in the 2015 report. However, the Committee considers the microbiome to be an emerging topic of potential importance to future DGACs.