FDNJ

Introduction

Pulses are the dry, edible seeds of legume plants, including dry beans, dry peas, chickpeas, and lentils. They do not include oilseed legumes like soybeans and peanuts, or legumes harvested fresh like green peas and green beans [1,2]. They are rich sources of protein, fiber, several important micronutrients and bioactive phytochemicals [3-6]. They are also a key component of nutritious, healthy and environmentally sustainable food systems [6,7] and are also a more affordable source of dietary protein [8]. The Dietary Guidelines for Americans (DGA) 2020-2025 recommends consuming 1.5 cups of beans, peas and lentils per week as part of healthy dietary pattern and includes them in both the vegetable group and the protein food group as their nutritional profile is similar to foods in both food groups [9]. The recently released 2025-2030 Dietary Guidelines Advisory Committee (DGAC) report recommended to increase intake of beans, peas and lentils up to 3 cups per week based on their modeling analysis [10]. Choose MyPlate of the United States Department of Agriculture (USDA) [11] considers dry beans, peas, and lentils as a part of protein food and ¼ cup cooked counts as one-ounce equivalent protein food while ½ cup prepared beans, peas and lentils is ½ cup equivalent of vegetables. Most dietary guidelines from around the world also recommend frequent consumption of pulses, mostly as part of protein foods [12,13].

Earlier analyses of National Health and Nutrition Examination Survey (NHANES) 1999-2002 data, indicated that approximately 27% of adult Americans consumed pulses with a mean intake of about 118 g/day over a two-day period [14]. Adult consumers of pulses were found to have higher intakes of fiber, protein, folate, zinc, iron and magnesium in analysis of NHANES 1999-2002 [14], and higher intakes of potassium, zinc, iron and choline in an analysis of NHANES 2003-2014 [15] compared to nonconsumers. However, these studies did not report association of pulse consumption on diet quality or prevalence of nutrient adequacy. In a dietary modeling analysis, we recently reported that addition of pulses replacing protein foods, refined grains, or combinations of protein foods and refined grains increased dietary fiber, iron, magnesium, potassium, and copper in the 2000 kcal Healthy U.S.-Style Pattern [8]. However, in that modeling analysis, we used USDA’s healthy dietary patterns that identify the amounts of foods in nutrient dense forms from all five major food groups to help people plan healthier meals but did not use the actual intakes and did not measure the effect of increasing pulses on diet quality or prevalence of nutrient adequacy. Regular intake of pulses has been associated with reducing the risk of chronic diseases such as cardiovascular diseases [16-21], cancer [22,23], diabetes [24,25], and overall mortality [26] in population studies and controlled trials. Dietary intake of pulses has also been found to be associated with higher satiety, lower food intake, weight loss and a reduction in body fat in randomized controlled trials [27-29].

Given this background, we hypothesized that pulses as nutrient rich source of protein and other nutrients would be associated with higher diet quality, nutrient intake, nutrient adequacy and increasing their consumption would further improve the diet quality and nutrient intake/adequacy. Therefore, the purpose of this study was to provide an updated evaluation of the association of pulses intake on diet quality, nutrient intake, and nutrient adequacy and to model the nutritional effects of increasing pulses in the current diets of adults using NHANES 2011–2018 dietary intake data [30].

Materials and Methods

Database: Dietary recall data from adults age 19+ years participating in the NHANES 2011-2018 (n = 17,234) cross-sectional survey were used to assess current intakes and association of intake of pulses with nutrient intake, nutrient adequacy and diet quality [30]. Data considered unreliable or incomplete by NHANES staff, and the data from pregnant or lactating women and from subjects with zero calorie on either day of recall, were excluded from the analysis. Demographic information and physical activity levels were determined from the NHANES 2011-2018 questionnaire data. Body mass index (BMI) and weight status as overweight (≥25 kg/m²) or obese (≥30 kg/m²) were determined from body measure files. All participants provided a signed written informed consent and the NHANES survey protocol was approved by the National Center for Health Statistics Ethics Review Board. The present study was a secondary data analysis that lacked personal identifiers and therefore was exempt from additional approvals by institutional review boards. A detailed description of the NHANES study protocol including subject recruitment, survey design, and data collection procedures as well as the data used in this study are publicly available and available online [30].

Dietary intakes, diet quality, and the % below the Estimated Average Requirement (EAR)/above the Adequate Intake (AI) in consumers and non-consumers of pulses: Intakes of food groups, energy and nutrients, and the % below the EAR/above the AI in consumers and non-consumers of pulses were evaluated using NHANES 2011-2018 24-hour dietary recall data using the relevant NHANES cycle of the USDA Food Patterns Equivalents Database (FPED) for food group intakes and the USDA Nutrient Database for Standard Reference Releases in conjunction with the respective Food and Nutrient Database for Dietary Studies for energy and nutrient intakes [31,32]. A total 341 food codes containing pulses were identified as determined using the FPED variable for measuring beans, peas and lentils (i.e., v_legumes > 0). Pulse consumers were defined as those individuals consuming any amount (cup eq) of pulses on day 1 or day 2 of the dietary recalls. Diet quality scores were determined using the USDA Healthy Eating Index (HEI) 2020 using the average of two days dietary intake data [33]. HEI 2020 has 13 components (8 adequacy components and 4 moderation components) each reflects key recommendations of DGA. Usual intake of nutrients and the % below the EAR/above the AI was determined with the National Cancer Institute (NCI) method (version 2.1) [34] using two days of dietary recall. Covariates used for the usual intake estimates were day of the week of the 24-hour dietary recall (coded as weekend [Friday to Sunday] or weekday [Monday to Thursday]) and sequence of dietary recall (first or second). The percent of the population below the EAR or above the AI was estimated using the cut-point method, except for iron, where the probability method was used [35].

Dietary modeling: First, all pulse foods reported in NHANES were identified. For each pulse item, energy and nutrient contents were extracted from the Food and Nutrient Database for Dietary Studies (FNDDS). A composite pulse profile was then created by calculating the sample-weighted mean energy and nutrient content across all pulse foods consumed in the population. This composite represented the nutrient composition of one “cup-equivalent” of pulses used in all modeling scenarios. Second, for every participant, we calculated the nutrient density of their reported diet by dividing each nutrient intake by total energy intake (e.g., g/kcal, mg/kcal, etc.). These nutrient-to-energy ratios were used to ensure that any calories removed from the diet (to maintain isocaloric conditions) reflected the participant’s own dietary pattern rather than removing arbitrary foods or nutrients. Third, for each modeling scenario, the amount of energy contributed by the added pulses was first calculated using the composite pulse profile. To maintain isocaloric intake, an equivalent amount of energy was removed from each participant’s baseline diet. Nutrients associated with the removed calories were calculated by multiplying the participant-specific nutrient-to-energy ratios by the number of calories removed. Thus, nutrients were removed proportionally to the participant’s existing diet composition. After subtracting these nutrients, the energy and nutrients from the modeled pulse intake were added to the participant’s diet. The following modeling scenarios were evaluated: 1) increase consumption of pulses to the level recommended in the DGA (1.5 cup eq/week), and 2) increase consumption of pulses to twice the level recommended in the DGA (3 cup eq/week).

To estimate usual intakes and the % below the EAR/above the AI associated with isocalorically adding pulses to the diet, usual intakes of nutrients was determined with the NCI method (version 2.1) [34] using two days of modified nutrient intakes. Covariates used in the usual intake estimates were day of the week of the 24-hour dietary recall (coded as weekend [Friday to Sunday] or weekday [Monday to Thursday]) and sequence of dietary recall (first or second). The percent of the population below the EAR or above the AI was estimated using the cut-point method, except for iron, where the probability method was used [35].

Statistical analyses: Analyses were performed using SAS 9.4 and all analyses were adjusted for the complex sampling (clustered sample) design of NHANES, using appropriate survey weights (oneday or two-day dietary sampling weight depending on analyses), strata, and primary sampling units. Regression analyses were used with to assess differences in consumers and non-consumers for demographics. Regression analyses were also utilized to assess differences (P<0.05) in energy and nutrient intakes, selected food group intakes, and diet quality after adjustment for age, gender, ethnicity, poverty income ratio levels, current smoking status, calories and BMI. Significant differences (P<0.05) for % below EAR/above AI between consumers and non-consumers were assessed via z statistics. Significant differences between baseline and modeling scenarios were assessed using non-overlapping 95^th percentile confidence intervals (CI) of means.

Results

Intake of pulses and consumer demographics: 34.9% adults age 19+ years were consumers of pulses with average intake of 1.41 ± 0.03 oz eq per day. Consumers of pulses were younger (-1.3 years), more likely to be male (3%), Hispanic (11%), Asian (2%), have vigorous physical activity (4%), have lower BMI (about 1 unit) and have normal weight (3% units ), and were less likely to be non-Hispanic White (7%), non-Hispanic Black (5%), sedentary (3%) and obese (5%) (Table 1).

Table 1: Demographics of adult (age 19+ years, gender combined, n = 17,234) consumers and non-consumers of pulses, NHANES 2011-2018 data.

Association of intake of pulses on intakes of energy, nutrients, food groups and diet quality: Adult consumers of pulses consumed 13% less added sugar than non-consumers (Table 2). Intakes of “total legumes”, “total fruit”, “total grain” and “total vegetables” were higher while intakes of “added sugars” and “beef, poultry & seafood” were lower among adult consumers of pulses than non-consumers (Table 2). Compared to nonconsumers, adult pulse consumers consumed 10% or more of dietary fiber, copper, iron, magnesium, potassium and folate. Adult pulse consumers had 4% more protein and 5% more sodium than non-consumers. Adult consumers of pulses had 13% higher HEI

2020 total scores non-consumers (Table 3). Component scores for adequacy components: “total vegetables”, “greens and beans”, “total fruit”, whole fruit”, “whole grains”, “total protein foods”, seafood and plant protein”, “fatty acid ratio” were higher for consumers indicating higher intakes compared to non-consumers. Components scores for moderation components: “saturated fat” and “added sugar” were also higher for consumers indicating lower intakes compared to non-consumers. However, component scores for moderation components: “sodium” and “refined grains” were lower for consumers indicating higher intake than non-consumers (Table 3).

Table 2: Intake of food groups, energy and nutrients among adult (age 19+ years, gender combined, n = 15,719) consumers and nonconsumers of pulses, NHANES 2011-2018 data.

Table 3: Healthy Eating Index (HEI) 2020 scores among adult (age 19+ years, gender combined, n = 15,719) consumers and nonconsumers of pulses, NHANES 2011-2018 data.

Association of intake of pulses on the % below the EAR/above the AI: A significantly lower proportion of adult consumers of pulses were below the EAR for magnesium (-27% units), for calcium, copper, zinc folate and vitamin C (-10% units or more), and for vitamin A, thiamin, vitamin B₆ and vitamin E (-5% or more) compared to non-consumers. Lower proportion of consumers compared to non-consumers were also below the EAR for other nutrients but the difference, while statistically significant, was less than -5% units). A higher  proportion of adult consumers of pulses were above the AI for dietary fiber (+14% units), potassium (+21% units) and choline (+5% units) as compared to non-consumers (Table 4).

Table 4: Nutrient adequacy among adult (age 19+ years, gender combined, n = 17,234) consumers and non-consumers of pulses, NHANES 2011-2018 data.

Modeling the effect of isocaloric addition of pulses on diet quality: Modeling increasing pulse consumption to the DGA recommended level (1.5 cup eq/week) and twice the recommended levels (3.0 cup eq/week) resulted in an increase in HEI   2020 total score by 8 and 13%, respectively (Table 5). Subcomponent scores also increased 9% and 18% for total vegetables, 118% and 187% for greens and beans, 6% and 11% for total protein foods, and

66% and 99% for seafood and plant protein by increasing pulse consumption to 1.5 and 3.0 cup eq/week, respectively. And since our modeling was conducted on an isocaloric basis, increasing pulse consumption to 3.0 cup eq/week additionally improved subcomponent scores by 5% for refined grains and by 6% for saturated fat (Table 5).

Table 5:  Modeling increasing pulses consumption to the recommended level and also twice the recommended levels (1.5 and 3.0 cup eq) in the DGA on diet quality (HEI 2020 scores) among adult (age 19+ years, gender combined, n = 17,234), NHANES 2011-2018 data.

Modeling the effect of isocaloric addition of pulses on nutrient With isocaloric addition of pulses to 3.0 cup eq, intakes of dietary intakes: Intake of dietary fiber (+12%), iron (+3%) and folate fiber (+23%), copper (+7%), iron (+6%), magnesium (+6%), (+5%) increased with isocaloric addition of pulses to 1.5 cup eq. potassium (+5%) and folate (+9%) increased (Table 6).

Table 6: Modeling increasing pulses consumption to the recommended level and also twice the recommended levels (1.5 and 3.0 cup eq/ week) in the DGA on nutrient intake among adult (age 19+ years, gender combined, n = 17,234), NHANES 2011-2018 data.