Part D. Chapter 2: Dietary Patterns, Foods and Nutrients, and Health Outcomes - Continued
For the first time, the 2015 DGAC included a chapter focusing solely on the relationship between dietary patterns and health outcomes. Although the 2010 DGAC considered some research on certain dietary patterns and specific health outcomes, notably body weight, they did not complete NEL systematic reviews on this research. The 2015 DGAC began by acknowledging a desire to continue and expand on the total diet approach initiated by the 2010 DGAC. They then identified outcomes of public health concern on which to focus their reviews.
For the purposes of the 2015 DGAC, dietary patterns were defined as the quantities, proportions, variety or combinations of different foods and beverages in diets, and the frequency with which they are habitually consumed. Because the purpose of the Dietary Guidelines is to develop food-based recommendations to promote health and reduce risk of diet-related disease, one of the key aspects of the research that the DGAC considered was a description of the foods and beverages consumed by participants in the studies that the Committee reviewed. This was particularly important for the NEL systematic reviews, for which a description of foods and beverages was a key criterion for inclusion. Data on nutrients were not required for inclusion, but were considered when provided as part of the dietary pattern description.
Self-reported food and beverage intake was typically assessed using a qualitative or semi-quantitative food intake questionnaire (i.e., FFQ). However, some studies used other methods, such as 24-hour recalls. When reviewing the evidence, the Committee attempted to adhere to the language used by the study authors in describing food groupings. There was variability across the food groupings, and this was particularly apparent in the meat group; for example, “total meat” may have been defined as “meat, sausage, fish, and eggs,” “red meat, processed meat, and poultry,” or various other combinations of meat. Similarly, “vegetables” seemed to most often exclude potatoes, but some studies included potatoes, yet they rarely provided information on how the potatoes were consumed (e.g., fried versus baked). When reported in the studies, the Committee considered these definitions in their review.
Because of the variability in dietary patterns methodology and food groupings reported, the Committee focused on providing a qualitative description of healthy dietary patterns. Additionally, as most studies reported intake in relative terms (e.g., comparing the first and fifth quintiles or across tertiles), the Committee has presented its conclusions with relative terminology (e.g., “higher” and “lower” in a certain component). Quantitative information on dietary patterns is provided in Part D. Chapter 1: Food and Nutrient Intakes, and Health: Current Status and Trends as part of the Dietary Patterns Composition section.
A number of studies in the scientific literature describe diets based on macronutrient proportion or test only a specific food group or nutrient in the diet. For example, a low-carbohydrate diet fits this description and has been of public interest. The DGAC reviewed the body of evidence related to this type of diet as part of Question 2. Additionally, the Committee examined the results of exploratory searches on low-carbohydrate diets (defined as less than 45 percent of calories from carbohydrate) and all of the health outcomes considered in this chapter published since 2000. Overall, it appears that only limited evidence is available to address the relationship between low-carbohydrate diets and health, particularly evidence derived from U.S.-based populations. The most evidence available focuses on low-carbohydrate diets and body weight. The 2010 DGAC examined the relationship between macronutrient proportion and various body weight outcomes, concluding that:
“1) There is strong and consistent evidence that when calorie intake is controlled, macronutrient proportion of the diet is not related to losing weight; 2) A moderate body of evidence provides no data to suggest that any one macronutrient is more effective than any other for avoiding weight re-gain in weight reduced persons; 3) A moderate body of evidence demonstrates that diets with less than 45% of calories as carbohydrates are not more successful for long-term weight loss (12 months). There is also some evidence that they may be less safe. In shorter-term studies, low-calorie, high-protein diets may result in greater weight loss, but these differences are not sustained over time; and 4) A moderate amount of evidence demonstrates that intake of dietary patterns with less than 45% calories from carbohydrate or more than 35% calories from protein are not more effective than other diets for weight loss or weight maintenance, are difficult to maintain over the long term, and may be less safe.”
The published literature since that review does not provide sufficient evidence to change these conclusions. Thus, in summary, although studies that examine macronutrient proportion or that test only a specific food group or nutrient are important, they answer different questions related to diet and health than those proposed by the DGAC. In addition, these studies generally did not meet the DGACs definition of a dietary pattern study unless a full description of the dietary pattern consumed was provided and appropriate methods were used to adjust for the confounding of foods and nutrients.
Questions 1, 2, and 3 were answered using existing reports, systematic reviews, and meta-analyses. All three of these questions were addressed in the NEL Dietary Patterns Systematic Review Project. This project was supported by USDAs Center for Nutrition Policy and Promotion and was informed by a Technical Expert Collaborative of experts in dietary patterns research.2 Additionally, the DGAC reviewed reports from systematic reviews recently conducted by the National Heart, Lung, and Blood Institute (NHLBI) that included dietary patterns research. For Question 1, the DGAC used the NHLBI Lifestyle Interventions to Reduce Cardiovascular Risk: Systematic Evidence Review from the Lifestyle Work Group10 and the associated American Heart Association (AHA)/ American College of Cardiology (ACC) Guideline on Lifestyle Management to Reduce Cardiovascular Risk.11 For Question 2, the DGAC used the NHLBI Managing Overweight and Obesity in Adults: Systematic Evidence Review from the Obesity Expert Panel12 and the associated AHA/ACC/The Obesity Society (TOS) Guideline for the Management of Overweight and Obesity in Adults.13 For all three questions, in an attempt to capture new research published since the searches for these systematic reviews were completed, the Committee considered existing systematic reviews and meta-analyses published in peer-reviewed journals since 2008. The existing systematic reviews and meta-analyses considered by the DGAC had to meet the general inclusion criteria of the DGAC, and were required to consider dietary patterns and the outcomes of interest. A description of the process the DGAC used to answer existing report questions is provided in Part C: Methodology. The DGAC followed this approach, including consideration of reference overlap, for all three questions. For more information on the existing reports, systematic reviews, and meta-analyses considered by the DGAC, the reader is encouraged to review the original sources, which are referenced within each evidence review.
Questions 4, 5, 6, and 7 were answered using NEL systematic reviews. A description of the NEL process is provided in Part C: Methodology. All reviews were conducted in accordance with NEL methodology, and the DGAC made all substantive decisions required throughout the process to ensure that the most complete and relevant body of evidence was identified and evaluated to answer each question. All steps in the process were documented to ensure transparency and reproducibility. Specific information about individual systematic reviews can be found at www.NEL.gov, including the search strategy, inclusion and exclusion criteria, a complete list of included and excluded articles, and a detailed write-up describing the included studies and the body of evidence. A link for each question is provided following each evidence review.
Introductory sections were written for Questions 4, 5, 6, and 7 because the conclusion statements for these questions were graded limited or insufficient. The purpose of the introduction was to provide a brief description of the current evidence available related to foods and nutrients and the health outcome of interest. However, this evidence was not considered in developing the dietary pattern conclusion statements. During the course of the dietary pattern reviews, the DGAC chose to highlight particular components of the diet, which are discussed further in Part D. Chapter 6: Cross-Cutting Topics of Public Health Importance.
Question 1: What is the relationship between dietary patterns and risk of cardiovascular disease?
Source of evidence: Existing reports
The DGAC concurs with the conclusions of the NEL Dietary Patterns Systematic Review Project and AHA/ACC Guideline on Lifestyle Management to Reduce Cardiovascular Risk that strong and consistent evidence demonstrates that dietary patterns associated with decreased risk of CVD are characterized by higher consumption of vegetables, fruits, whole grains, low-fat dairy, and seafood, and lower consumption of red and processed meat, and lower intakes of refined grains, and sugar-sweetened foods and beverages relative to less healthy patterns. Regular consumption of nuts and legumes and moderate consumption of alcohol also are shown to be components of a beneficial dietary pattern in most studies. Randomized dietary intervention studies have demonstrated that healthy dietary patterns exert clinically meaningful impact on cardiovascular risk factors, including blood lipids and blood pressure. Additionally, research that includes specific nutrients in their description of dietary patterns indicate that patterns that are lower in saturated fat, cholesterol, and sodium and richer in fiber, potassium, and unsaturated fats are beneficial for reducing cardiovascular disease risk. DGAC Grade: Strong
Individuals are encouraged to consume dietary patterns that emphasize vegetables, fruits, whole grains, legumes, and nuts; include low-fat dairy products and seafood; limit sodium, saturated fat, refined grains, and sugar-sweetened foods and beverages; and are lower in red and processed meats. Multiple dietary patterns can achieve these food and nutrient patterns and are beneficial for cardiovascular health, and they should be tailored to individuals biological needs and cultural as well as individual food preferences. The Committee recommends the development and implementation of programs and services at the individual and population levels that facilitate the improvement in eating behaviors consistent with the above dietary patterns.
Review of the Evidence
The DGAC examined research compiled in the NEL Dietary Patterns Systematic Review Project, which included 55 articles summarizing evidence from 52 prospective cohort studies and 7 RCTs, and the 2013 AHA/ACC Lifestyle Guideline and associated NHLBI Lifestyle Report, which included primarily RCTs. The Committee drew additional evidence and effect size estimates from six published systematic reviews/meta-analyses published since 2008 that included one or more studies not covered in the NEL or NHLBI Lifestyle reports.14-19 In total, 142 articles were considered in these reports, of which 35 were included in two or more reviews. Little evidence on the contribution of dietary patterns to CVD risk factors in the pediatric populations was available, and that which was published was not systematically reviewed.
Most evidence examining hard disease endpoints comes from large, prospective cohort studies in adults using a priori scores to rank individuals with respect to adherence to dietary patterns of interest. Though the observational design allows the necessary duration of follow-up to observe CVD endpoints, comparison across studies was difficult because of different methods for deriving scores and different versions of scores measuring adherence to the same dietary pattern. In the Mediterranean dietary indices and the AHEI scores, moderate alcohol was included as a “positive” component (associated with potential benefits). Red and processed meats were “negative” (potentially detrimental) components in the Mediterranean scores, AHEI scores, and DASH. Certain scores also included sugars or sugar-sweetened beverages as negative components. Poultry was considered as a positive component in the original AHEI. Total high-fat dairy was a negative component in the Mediterranean diet scores, but dairy was a positive component when meeting recommended intakes for the HEI-2005, and low-fat dairy was positive in the DASH scores. As the NEL systematic review points out, several components of scores associated with decreased CVD risk recurred in multiple dietary patterns and were associated as part of scores and as individual components with reduced CVD risk. These included consumption of vegetables, fruits, whole grains, nuts, legumes, unsaturated fats, and fish.
The NHLBI Lifestyle Report summarized the evidence from two RCTs of the DASH dietary pattern and two trials testing DASH variations with differing levels of sodium or macronutrients. The diet provided to participants in standard DASH intervention trials was high in vegetables, fruits, low-fat dairy products, whole grains, poultry, fish, and nuts. It also was low in sweets, sugar-sweetened beverages, and reduced in (or lower in) red and processed meats. The DASH dietary pattern is high in fiber and potassium and low in sodium, saturated fat, total fat, and cholesterol. It is rich in potassium, magnesium, and calcium, as well as protein and fiber.
In contrast to the patterns described above, vegetarian diets were defined by what they excluded. Variations included: vegan (no meat, fish, eggs, or dairy); lacto-ovo vegetarian (includes eggs and dairy, but no fish or meat), and pesco-vegetarian (includes fish, but no meat) diets. The content of these diets varied substantially, though they tended to emphasize plant based foods, especially fruits and vegetables, legumes, nuts, and whole grains.
Dietary Patterns and Blood Pressure (BP)
DASH or DASH-style Dietary Patterns
The NEL systematic review and AHA/ACC Lifestyle Guideline conclude that strong and consistent evidence from RCTs demonstrates that compared to a dietary pattern that is relatively high in saturated fat and sodium and low in vegetables and fruits, the DASH-style dietary pattern reduced BP by approximately 6/3 mmHg (systolic blood pressure/diastolic blood pressure) across subgroups defined by sex, race, age, and hypertension status. The DASH trial provided all food to participants for 8 weeks. Fat intake was relatively low at 26 percent of energy (7 percent each monounsaturated and saturated, 10 percent polyunsaturated), compared to 36 percent in the control group. Carbohydrates accounted for 57 percent of energy and protein for 18 percent. Sodium was stable at 3,000 mg/day and body weight did not change. Variations of the DASH diet also lowered blood pressure: in the OmniHeart Trial, compared to the standard DASH, replacing 10 percent of calories from carbohydrate with either the same calorie content of protein or with unsaturated fat (8 percent MUFA and 2 percent PUFA) lowered systolic BP by 1 mmHg. Among adults with BP 140159/9095 mmHg, these substitutions lowered systolic BP by 3 mmHg relative to standard DASH.2 11
Observational evidence summarized in the NEL report included one cohort showing that increased DASH score was associated with small, but decreased levels of systolic and diastolic BP over time;20 two others cohorts showed no relationship between DASH scores and risk of hypertension.21 22
Mediterranean-Style Dietary Patterns
Several RCTs provide limited to moderate evidence on the benefits of a Mediterranean-style diet for reducing blood pressure. The AHA/ACC Lifestyle Guideline conclude that consuming a Mediterranean dietary pattern instead of a lower-fat dietary pattern had beneficial effects on blood pressure. The NHLBI Lifestyle Report reviewed two RCTs of free-living middle-aged or older adults (with type 2 diabetes or at least three CVD risk factors) in which a Mediterranean diet intervention reduced BP by 67/23 mmHg.23 24 The report also reviewed one observational study of healthy younger adults. Higher adherence to a Mediterranean-style diet, as measured through a Mediterranean score, was associated with a decrease in BP of 23/12 mmHg.25
Vegetarian Dietary Patterns
Evidence for the blood pressure benefits of vegetarian dietary patterns is more limited, but moderately consistent trends appear to exist. A recent meta-analysis of seven RCTs found that consumption of vegetarian diets was associated with a reduction in mean systolic blood pressure (-4.8 mm Hg; 95% CI = -6.6 to -3.1; p<0.01) and diastolic blood pressure (-2.2 mm Hg; 95% CI = -3.5 to -1.0) compared with the consumption of omnivorous diets.19 The AHA/ACC Lifestyle Guideline did not find sufficient evidence to examine vegetarian dietary patterns, and the NEL systematic review summarized only three studies comparing blood pressure outcomes in lacto-ovo vegetarian diets versus non-vegetarian diets in which meat and fish were consumed. Of the two studies, one was a large prospective cohort that found no association with blood pressure,26 and the other was a RCT among individuals with hypertension that demonstrated a decrease in systolic blood pressure, but not diastolic blood pressure.27 The more recent EPIC-Oxford cohort found lower systolic, but not diastolic blood pressure compared to the findings of Crowe, 2013.28
Other Dietary Patterns
As summarized in the NEL systematic review, adherence to the 2005 Dietary Guidelines for Americans was related to lower blood pressure in one study of healthy young adults. Zamora et al reported 20-year findings from the CARDIA study including 4,381 Black and White young adults.29 Participants in the highest (vs. lowest) quartile of adherence to the 2005 Dietary Guidelines had significantly less increase in systolic and diastolic blood pressure over time.
Dietary Patterns and Blood Lipids
DASH or DASH-style Dietary Patterns
As reviewed in the NHLBI Lifestyle Report, RCTs of the DASH diet show favorable effects on low-density lipoprotein cholesterol (LDL-C) and total cholesterol: high-density lipoprotein cholesterol (total-C: HDL-C) ratio, and no effect on triglycerides (TG). Benefits were seen with a variety of different macronutrient compositions, though they were enhanced when some carbohydrates in the standard DASH pattern were replaced with protein or unsaturated fat. In the standard DASH, when food was supplied to adults with a total cholesterol level of less than 260 mg/dL and LDL-C less than 160 mg/dL, and body weight was kept stable, the DASH dietary pattern compared to the control diet decreased LDL-C by 11 mg/dL, decreased HDL-C by 4 mg/dL, and had no effect on TG. The OmniHeart trial tested the DASH dietary pattern with different macronutrient compositions among adults with average baseline LDL-C 130 mg/dL, HDL-C 50 mg/dL, and TG 100 mg/dL. Modifying the DASH diet by replacing 10 percent of calories from carbohydrate with 10 percent of calories from protein decreased LDL-C by 3 mg/dL, decreased HDL-C by 1 mg/dL, and decreased TG by 16 mg/dL compared to the DASH dietary pattern. Replacing 10 percent of calories from carbohydrate with 10 percent of calories from unsaturated fat (8 percent MUFA and 2 percent PUFA) decreased LDL-C similarly, increased HDL-C by 1 mg/dL, and decreased TG by 10 mg/dL compared to the DASH dietary pattern.11
Mediterranean-style Dietary Patterns
As with blood pressure, few trials have evaluated the effects of Mediterranean dietary patterns on blood lipids. According to the AHA/ACC Lifestyle Guideline, consuming a Mediterranean-style diet (compared to minimal or no dietary advice) resulted in no consistent effect on plasma LDL-C, HDL-C, and TG. In part, this was due to substantial differences in dietary interventions conducted among free-living middle aged or older adults with or without CVD or at high risk for CVD.11 In the PREDIMED trial (reviewed in both the NHLBI Lifestyle and NEL reports), both treatment groups (Mediterranean diet +olive oil or +nuts) had favorable changes in HDL-C, total-C: HDL-C ratio and TG when compared to the control group, which received minimal advice to follow a lower-fat diet.23 One of the prospective cohort studies reviewed by the NEL showed each one-point increase in alternate Mediterranean diet score assessed in adolescence and early adulthood was associated with a -6.19 (-10.44, -1.55) mg/dL lower total cholesterol in adulthood but no significant effects on HDL-C.30 Of other observational cohorts reviewed, one reported adherence to a Mediterranean diet was associated with favorable changes in HDL-C and TG,31 and another found no associations between adherence to a Mediterranean diet and blood lipids.32
Vegetarian Dietary Patterns
The NEL systematic review included three articles on vegetarian patterns that measured blood pressure or blood lipids.26-28 One study reported decreased total-C26 and another reported decreased non-HDL-C in vegetarian versus non-vegetarian participants.28
Other Dietary Patterns
Of note, adherence to the 2005 Dietary Guidelines for Americans also was related to higher HDL-C levels in a cohort of Black and White young adults.29
Dietary Patterns and Cardiovascular Disease Outcomes
The NHLBI Lifestyle review did not include any trials examining the evidence of particular dietary patterns with CVD outcomes. Overall, the NEL systematic review found that individuals whose diets mirrored the dietary patterns of interest (typically compared with diets having lower scores) was associated with lower CVD incidence and mortality in 14 out of 17 studies. The studies were predominantly observational, but included some trial evidence, and they typically assessed dietary intakes through self-report. The effect sizes varied substantially, with the decrease in risk of CVD ranging from 22 to 59 percent for increased adherence to various Mediterranean-style dietary patterns and from 20 to 44 percent for increased adherence to a U.S. Dietary Guidelines-related pattern (e.g., HEI or AHEI and updates). The majority of studies that assessed coronary heart disease (CHD) incidence or mortality also reported a favorable association between adherence to a healthy dietary pattern and CHD risk. The lower CHD risk ranged from 29 to 61 percent for greater adherence to Mediterranean-style dietary patterns, from 24 to 31 percent for greater adherence to a U.S. Dietary Guidelines-related pattern, and from 14 to 27 percent for greater adherence to DASH. Similarly, the majority of studies assessing stroke incidence or mortality reported favorable associations, with the lower stroke risk ranging from 13 to 53 percent for greater adherence to a Mediterranean-style dietary pattern and from 14 to 60 percent for greater adherence to a U.S. Dietary Guidelines-related pattern.2
Mediterranean-style Dietary Patterns
To gather additional information on dietary patterns and CVD outcomes, the DGAC consulted two meta-analyses,15 18 which included many of the same observational prospective cohort studies as one another and as the NEL systematic review. These meta-analyses each reported summary estimates across studies as a 10 percent reduction in risk of CVD (fatal or nonfatal clinical CVD event) per 2-increment increase in adherence to the Mediterranean-style diet. The NEL report also included results from the largest Mediterranean diet trial, PREDIMED, which found that a Mediterranean diet (plus extra virgin olive oil or nuts) had favorable effects in high-risk participants compared to the control group who were advised to reduce dietary fat intake. An approximately 30 percent decrease in risk of major CVD events (a composite endpoint including myocardial infarction, stroke, and deaths) was observed and the trial was stopped early for meeting benefit requirements.2 33 According to food questionnaires measuring adherence to the assigned diet by the end of follow-up, the intervention groups had significantly increased consumption of fish and legumes and non-significant reductions in refined grains and red meat from baseline, in addition to increased intake of supplemental foods (olive oil or nuts depending on the intervention arm), compared to the control group.
DASH-style Dietary Patterns
A recent meta-analysis17 of six prospective cohort studies with CVD endpoints assessed DASH-style diet through the Fung et al. method,6 which assigns points based on population-specific quintiles of eight DASH dietary pattern components: fruits, vegetables, nuts and legumes, whole grains, low-fat dairy, sodium, red and processed meats, and sweetened beverages. This meta-analysis reported that greater adherence to a DASH-style diet significantly reduced CVD (Relative Risk [RR]=0.80; 95% CI = 0.74 to 0.86), CHD (RR=0.79; 95% CI = 0.71 to 0.88), and stroke (RR=0.81; 95% CI = 0.72 to 0.92). All of the studies meta-analyzed also were included the NELs evidence base for the DASH-style diet.
Vegetarian Dietary Patterns
The NEL systematic review concluded that evidence for the effects of vegetarian dietary patterns on cardiovascular endpoints is limited. Most of this evidence was from prospective cohort studies; four out of six studies suggested that a vegetarian dietary pattern was associated with reduced incidence of ischemic heart disease (IHD) or CVD mortality. A meta-analysis of seven studies related to CVD mortality and vegetarian diet14 (including two of the studies from the NEL systematic review) found that mortality from IHD was significantly lower in vegetarians than in non-vegetarians (RR=0.71; 95% CI = 0.56 to 0.87). The authors estimated a 16 percent lower mortality from circulatory diseases (RR=0.84; 95% CI = 0.54 to 1.14) and a 12 percent lower mortality from cerebrovascular disease (RR=0.88; 95% CI = 0.70 to 1.06) in vegetarians compared to non-vegetarians.
For additional details on this body of evidence, visit: References 2, 10, 11, 14-19 and Appendix E-2.26