Background:
This important public health question was identified by the U.S. Departments of Agriculture (USDA) and Health and Human Services (HHS) to be examined by the 2020 Dietary Guidelines Advisory Committee.
The 2020 Dietary Guidelines Advisory Committee, Dietary Patterns Subcommittee conducted a systematic review to answer this question with support from the Nutrition Evidence Systematic Review (NESR) team.
The goal of this systematic review was to examine the following question: What is the relationship between dietary patterns consumed and all-cause mortality?
Conclusion statements and grades:
Dietary patterns
Strong evidence demonstrates that dietary patterns in adults and older adults characterized by vegetables, fruits, legumes, nuts, whole grains, unsaturated vegetable oils, and fish, lean meat or poultry when meat was included, are associated with decreased risk of all-cause mortality. These patterns were also relatively low in red and processed meat, high-fat dairy, and refined carbohydrates or sweets. Some of these dietary patterns also included alcoholic beverages in moderation. (Grade: Strong)
Diets based on macronutrient distribution
Insufficient evidence was available to determine the relationship between diets based on macronutrient distribution and all-cause mortality. (Grade: Grade not assignable)
Methods:
A literature search was conducted using 3 databases (PubMed, Cochrane, Embase) to identify articles that evaluated the intervention or exposure of dietary patterns consumed and the outcome of all-cause mortality. A manual search was conducted to identify articles that may not have been included in the electronic databases searched. Articles were screened by two NESR analysts independently for inclusion based on pre-determined criteria
Data extraction and risk of bias assessment were conducted for each included study, and both were checked for accuracy. The Committee qualitatively synthesized the body of evidence to inform development of a conclusion statements, and graded the strength of evidence using pre-established criteria for risk of bias, consistency, directness, precision, and generalizability.
Dietary patterns were defined as the quantities, proportions, variety, or combination of different foods, drinks, and nutrients (when available) in diets, and the frequency with which they are habitually consumed.
Diets based on macronutrient distribution were examined when at least one macronutrient proportion was outside of the acceptable macronutrient distribution range (AMDR) for carbohydrate, fat, and/or protein, whether or not the foods/food groups consumed were provided
Summary of the evidence:
This systematic review identified 153 articles, including one randomized controlled trial and 152 prospective cohort study designs that met inclusion criteria and were published between January 2000 and October 2019.
141 studies examined the relationship between dietary patterns and all-cause mortality. The studies used multiple approaches to assess dietary patterns and all-cause mortality.
One RCT assigned participants to consume a “Mediterranean” dietary pattern with extra-virgin olive oil or mixed nuts compared to a control diet
One-hundred ten articles examined dietary patterns using index or score analysis,
Twenty-five articles examined dietary patterns identified with factor and cluster analysis,
Eleven articles used other methods, including only reduced rank regression (RRR), comparisons based on animal-product consumption vs. avoidance, or comparisons based on ‘ultra-processed’ food consumption, to examine the relationship between dietary patterns and/or diets based on macronutrient distribution
Despite the variety of different methods applied to examine or derive dietary patterns, the majority of studies finding statistically significant relationships between dietary patterns consumed and all-cause mortality risk were remarkably consistent.
Although the dietary patterns examined were characterized by different combinations of foods and beverages due to the variety of methods used, protective dietary patterns emerged with the following themes:
Patterns emphasizing higher consumption of vegetables, legumes, fruit, nuts, whole grains, fish, lean meat or poultry, and [unsaturated fats relative to saturated fats (either as a ratio of polyunsaturated fatty acids (PUFA) and monounsaturated fatty acids (MUFA) relative to saturated fats/fatty acids (SFA), MUFA relative to SFA, or Olive Oil specifically] were generally associated with decreased risk of all-cause mortality. Notably, there was consistency in particular with the inclusion of fish and/or seafood. Some of these dietary patterns also included alcoholic beverages in moderation.
Reduced risk of all-cause mortality was observed in several studies that examined dietary patterns without animal-source foods, such as those described as vegetarian, vegan, or determined by “plant-based” diet indices.
Of the dietary patterns that included animal-source foods, protective associations were generally observed with relatively lower consumption of red and processed meat. However, a limitation in the evidence is methodological heterogeneity in the food categories and terminology used to classify meat.
The inclusion of the ratio of white vs. red meat, type and amount of dairy products, and refined carbohydrates and sweets as elements to these patterns was less consistent across the evidence. The dietary patterns that included those elements and tended to show reduced risk of all-cause mortality had,
higher consumption of white meat relative to red or processed meat,
low-fat dairy relative to high-fat dairy, and/or
lower relative to higher intake of refined carbohydrates and sweets.
Despite the variability between approaches used to examine dietary patterns, higher adherence to dietary patterns with common labels such as “Mediterranean”, dietary-guidelines related (e.g., “Healthy Eating Index”, “DASH” scores), or “plant-based” were generally protective against all-cause mortality risk. This highlights that a high-quality dietary pattern comprised of nutrient-dense foods, regardless of the label, associated with reduced all-cause mortality risk.
Results based on additional analyses according to a variety of key or potential confounders generally confirmed the robustness of results.
Although the majority of included studies were prospective cohort studies, most adjusted for key confounders, with the exception of race and ethnicity. The results are likely generalizable to adults of various race and ethnicity though it is difficult to determine the impact that race and ethnicity specifically may have on the relationship between dietary patterns and all-cause mortality due to a lack of reporting.
Insufficient evidence was available to determine the relationship between dietary patterns and all-cause mortality in younger populations (~age <35 years)
Twenty-eight articles examined the relationship between diets based on macronutrient distributions but results were inconsistent.
When describing and categorizing studies included in this review, the Committee did not label the diets examined as “low” or “high,” because no universally accepted standard definition is currently available for “low-carbohydrate” or “high-fat” diets. Instead, the Committee focused on whether, and the extent to which, the proportions of the macronutrients were below or above the AMDR.
Diets with proportions of carbohydrate and fat within the AMDR compared to outside the AMDRs tended to associate with reduced all-cause mortality risk, particularly when the diets examined were of higher quality (i.e., emphasizing vegetables, fruits, nuts, whole grains, legumes, fish, and/or lean meat or poultry).
Comparison of macronutrient distributions with or without the context of the foods/food groups comprising the dietary pattern showed inconsistent findings, likely due to several limitations that prevent the adequate assessment of the body of evidence:
The gradient between the macronutrient proportions compared between distributions was often small, e.g., 41% vs. 41.7%
Methods used to estimate macronutrient intake differed between studies
Many of the proportions outside of the AMDR were only marginally outside and often estimated differently between studies.
Most of these articles reported a proportion of energy from carbohydrate below and/or fat above the AMDR in at least one of the exposure groups compared.
Some of these articles also described the dietary pattern (i.e., foods and beverages) consumed, in addition to having macronutrient proportions outside of the AMDR.