Abstract
Background
Although poor diet quality is a main risk factor for cardiovascular disease, it is rarely measured comprehensively in clinical trials using full dietary assessment tools.
Objectives
This study developed a Dietary Approaches to Stop Hypertension (DASH) diet screener using questions from the Stenting vs. Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis trial (SAMMPRIS) and assessed construct validity in 3 2-y cycles of the National Health and Nutrition Examination Survey (2009–2014).
Methods
Two DASH scores were created in a subset of 14,651 adults aged ≥20 with ≥1 reliable 24-h diet recall and plausible energy intakes. The validated DASH score used established methods to create a score ranging from 8 to 40. The DASH screener used 11 nonvalidated SAMMPRIS questions to create 8 components. Each component was multiplied by a weight for comparability between the scores and summed to obtain an overall score (0–100), with higher values indicating better adherence. Construct validity was examined by analyzing whether the screener score had a variable distribution, correlated with the validated score, differentiated groups with known diet quality differences, and was concordant with the validated score.
Results
Participants were on average 47.6 (±16.9) y, 68% non-Hispanic White, 53% female, and with a body mass index (BMI) of 28.9 kg/m2. The mean (SD) of the DASH screener score and validated DASH score were 47.0 (14.7) and 23.8 (5.2). The elements of construct validity were demonstrated with strong correlations between the total and most components' scores (r = 0.62–1.00, P < 0.0001), the ability to distinguish known-group differences, and strong concordance between the 2 scores (κ = 0.62, P< 0.0001). In sensitivity analyses, removal of the sodium component improved the total score correlation (r = 0.73) and concordance (κ = 0.64).
Conclusions
The brief DASH diet screener demonstrated construct validity across 4 domains. Future research is needed to better estimate salt intake and evaluate the predictive validity of the screener.
Keywords: dietary screening, dietary pattern, Dietary Approaches to Stop Hypertension (DASH) diet, cardiovascular health, construct validity
Introduction
Cardiovascular diseases (CVD) continue to be the leading cause of death in the United States [1]. Poor diet is a primary risk factor for elevated risk of CVD mortality and morbidity, with 7.9 million annual deaths (14% of all deaths) and 188 million DALYs (7% of all disability-adjusted life-years) being attributable to dietary risk factors globally [2,3]. Greater adherence to heart-healthy dietary patterns, such as the Dietary Approaches to Stop Hypertension (DASH) diet, has been consistently linked with lower risk of CVD in epidemiological studies [[4], [5], [6]].
Recognizing the importance of healthy dietary patterns, numerous CVD clinical trials provide nutrition or diet counseling as a part of their interventions. However, due to the burdens of high costs, time constraints, and competing demands, dietary pattern adherence may only be assessed briefly using nonvalidated tools, especially when diet is not the primary outcome of interest [7,8]. The need to assess diet in a way that is feasible and balances these constraints has led to a call in the field for rapid diet screener tools that can quickly assess dietary habits and nutritional intake with a small set of questions [7]. Ideally, a valid rapid diet screener would reflect one’s overall diet quality similarly as those measured by full dietary assessment methods such as food frequency questionnaire (FFQ) or 24-h dietary recall, differentiate groups with known differences in diet quality and adherence, and have higher scores that are predictive of better health outcomes [7,9].
In the Stenting vs. Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis (SAMMPRIS) clinical trial (2008–2013) [10], all participants were enrolled in a lifestyle modification program as part of the intervention and received a nutritional education module that emphasized dietary practices that align strongly with the DASH-style dietary pattern [11]. Briefly, the DASH diet emphasizes intake of fruits and vegetables, fat-free or low-fat dairy, whole grains, and lean protein, and limits intake of red and processed meats, sweets, foods and beverages with added sugars, saturated fats, sodium, and alcohol. To capture adherence and intake of key DASH components, a set of nonvalidated questions was administered in lieu of a full dietary assessment. Although these questions capture elements of DASH dietary patterns, they differ from traditional recall methods in important ways, including not being linked to a nutrient database, and it is unclear how well they measure diet quality compared with established indices.
Evaluating whether these questions used in SAMMPRIS validly capture DASH diet quality is important to establish their utility in future clinical research. However, in the absence of a full dietary assessment method, it is not possible to validate these questions using reference standard methods. Nevertheless, examination of construct validity can be achieved by evaluating screener performance in datasets such as the NHANES, which collected dietary data using multiple 24-h dietary recalls. Therefore, this study aimed to use the set of nonvalidated SAMMPRIS questions to develop a DASH diet screener score following the component weighing strategies of an established DASH adherence score (Fung DASH score) [12], and to examine its construct validity in 3 2-y cycles of NHANES (2009–2014).
Methods
Study population and design
NHANES is a nationally representative, cross-sectional survey, conducted by the National Center for Health Statistics, of the noninstitutionalized United States population [13]. Approximately 5000 individuals of all ages are interviewed and complete the health examination component of the survey every year with data publicly released in 2-y cycles, since 1999 [13]. To evaluate construct validity, the DASH diet screener score was calculated and tested against the Fung DASH score in a cross-sectional analysis of the 2009–2014 cycles of the NHANES. These specific cycles were selected to best align with the time during which the SAMMPRIS trial was conducted. The study design and methods for the 2009-2014 cycles have been previously published. For dietary interview components, all NHANES examinees are eligible for 2 24-h dietary recall interviews. The first dietary recall is collected in-person in the Mobile Examination Center using the automated multiple-pass method and the second interview is collected by telephone 3–10 d later [14]. A subset of 14,651 adults aged 20 y and older who had ≥1 reliable 24-h dietary recall indicated by dietary recall status code and predefined plausible energy intake levels (800–4200 kcal/d for males and 500–3500 kcal/d for females) was created (Figure 1) [[15], [16], [17]]. The University of Rhode Island Institutional Review Board approved the study protocol.
FIGURE 1.
NHANES 2009–2014 participants flowchart. Illustration of the creation of the analytical sample (n = 14,651) from NHANES 2009–2014 (n = 30,468) after excluding participants who did not meet the study eligibility criteria.
Calculation of food and nutrient intakes based on DASH components
To calculate food and nutrient intakes for DASH components, each food and beverage in the Individual Food Files (IFF) generated from NHANES 24-h recalls was linked to the USDA Food and Nutrition Database for Dietary Studies (FNDDS) and USDA Food Patterns Equivalents Database (FPED) using 8-digit USDA food code. FPED converts the food and beverages in the FNDDS to the 37 USDA Food Pattern components measured as cup/ounce/gram equivalents per 100 g of each food [18]. DASH relevant food pattern components include intakes of total fruits, total vegetables, nuts and legumes, whole grains, seafood, low-fat dairy, and red and processed meat. Information on nutrients such as sodium and saturated fats was calculated from the Total Nutrients Intakes files from NHANES 24-h recalls. A variable for sugar-sweetened beverages (SSB) was created, and intakes were calculated from IFFs using FNDDS and FPED.
Calculation of Fung DASH score and the DASH diet screener score
The Fung DASH score was calculated using a previously published method [12]. Briefly, intake of each of the 8 Fung DASH components (Table 1) was calculated from 24-h recalls. For scoring, quintile ranking ranging from 1 to 5 was applied to each component, with higher scores assigned to higher quintiles of intake, except for sodium, red and processed meats, and SSB, where scoring was reversed. An overall DASH adherence score ranging from 8 to 40 was obtained by summing all the component scores. The DASH diet screener score was created using a set of questions from SAMMPRIS nutrition questionnaire, which highlighted 8 key components in the DASH diet: increasing the intake of fruits and vegetables, legumes, low-fat dairy, fish as lean protein, and limiting the intake of sodium, red meat, saturated fat, and SSB (Table 1). Although whole grain intake is also encouraged in the DASH diet, it was not included in the scoring because the screener did not include questions on whole grains; hence, its intake could not be estimated. For assigning weight across components and classifying intake by quintiles within each component when serving size information was provided, the scoring scheme of Fung DASH was used as the theoretical basis [12]. The underlying premise was that all components contribute equally to the total score, except for the component on fruits and vegetables, which was weighed twice, as their frequencies of intake were collected jointly in the questionnaire and could not be separated into distinct components. Higher scores were assigned to either higher frequencies of intake or higher quintiles of intake for the components encouraged by the DASH dietary pattern (Table 2). Scores were reverse-coded for components associated with elevating CVD risks and for which low consumption is desired, including sodium, red meat, saturated fat, and SSB. For sodium and saturated fat components, where 2 screening questions regarding the same construct were administered to capture frequency of intake, an average score of the 2 questions was taken to compute the corresponding component score. An overall DASH diet screener score, ranging from 0 to 100, was calculated by multiplying the screener component scores by their corresponding weights and summing the results, with higher scores indicating better DASH adherence and better diet quality. See Table 1 for the algebraic equation for the calculation of the total score.
TABLE 1.
Scoring criteria for the DASH-style diet in Fung DASH and brief DASH screener score.
| Fung DASH score components1 | Component score range | Weight | Brief DASH screener score components2 | Component score range | Component weight |
|---|---|---|---|---|---|
| Adequacy | |||||
| Fruits | 1–5 | 12.5% | Fruits + vegetables | 0–3 | 22% |
| Vegetables | 1–5 | 12.5% | |||
| Nuts and legumes | 1–5 | 12.5% | Legumes | 0–3 | 11% |
| Whole grains | 1–5 | 12.5% | Low-fat dairy | 0–3 | 11% |
| Low-fat dairy | 1–5 | 12.5% | Fish | 0–3 | 11% |
| Moderation | |||||
| Sodium | 1–5 | 12.5% | Sodium | 0–2 | 11% |
| Red processed meat | 1–5 | 12.5% | Red meat | 0–3 | 11% |
| SSB | 1–5 | 12.5% | Saturated fat | 0–3 | 11% |
| SSB | 0–4 quintile | 11% | |||
| Total score range | 8–40 | 0–100 | |||
Algebraic equation for the calculation of the brief DASH screener score:
| |||||
Abbreviations: DASH, Dietary Approaches to Stop Hypertension; SSB, sugar-sweetened beverages.
Fung DASH score was calculated by applying quintiles ranging from 1 to 5 to each component and summing all 8 components to get the total score, ranging from 8 to 40, with higher scores assigned to higher quintiles of intake for adequacy components and lower scores assigned to higher quintiles of intake for moderation components. A higher total Fung DASH score indicated better DASH adherence.
Brief DASH screener score was calculated by multiplying the screener component scores by their corresponding component weights and summing the results, with higher scores assigned to higher intake for adequacy components and lower scores assigned to higher quintiles of intake for moderation components. A higher total brief DASH screener score indicated better DASH adherence.
TABLE 2.
Screener questions for DASH diet component assessment and score assignment.
| Components | Screener questions | Thresholds |
|---|---|---|
| 1. Fruits + vegetables | How many servings of fruits and vegetables do you eat per day? |
|
| 2. Legumes | How often do you eat soy products/beans? |
|
| 3. Low-fat dairy | How often do you eat reduced-fat or nonfat dairy products? |
|
| 4. Fish | How often do you eat fish? |
|
| 5. Sodium | 1. How often do you add salt to your food after it is served to you? 2. If you cook, how often do you add salt in cooking or preparing food? |
|
| 6. Red meat | How often do you eat red meat? |
|
| 7. Sugar-sweetened beverages | How many of each of the following sugar-containing beverages do you consume in an average day? -Regular soft drinks/d -Sweetened iced tea/d |
|
| 8. Saturated fat | 1. How often do you eat high-fat snacks, chips, or fried food? 2. How often do you use butter at the table or in cooking? |
Question 1:
|
Abbreviation: DASH, Dietary Approaches to Stop Hypertension.
Groups with known differences in diet quality
Variables were selected a priori based on their known differences in diet quality, including sociodemographic information on age groups (20–39, 40–59, and 60+), sex (male and female), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, and other race, including multiracial), and education level (high school or less, some college, and college graduate or higher), anthropometric information including BMI (in kg/m2), and health behavioral information including smoking status and physical activity [9,19,20]. BMI categories were created and reported as underweight (<18.5), normal weight (18.5 to <25), overweight (25 to <30), and obese (≥30). For health behavioral risk factors, smoking status was reported as never, former, and current smokers. Never smokers were defined as those who had either never smoked or had smoked <100 cigarettes in their lifetime, former smokers as those who had smoked >100 cigarettes but no longer smoking, and current smokers as those who had smoked >100 cigarettes in their lifetime and were still smoking. Physical activity was assessed based on the Global Physical Activity Questionnaire [21], from which metabolic equivalents of task (MET) minutes of moderate and vigorous recreational activity per week were estimated. Physical activity status was categorized as sedentary (MET minute = 0), lightly active (0 < MET min < 600), and active (MET min ≥ 600), which was based on national guidelines [22].
Construct validity testing
To evaluate construct validity, the DASH screener score was examined to determine its ability 1) to show sufficient variation in the distribution of scores across the United States population, 2) to show correlation with an established DASH diet score, 3) to distinguish groups with known difference in diet quality (also referred to as known groups validity), and 4) to show concordance with an established DASH diet score. Strategies for construct validity testing, adopted from the American Heart Association (AHA) scientific statement on rapid dietary assessment and validation studies for other diet quality measures, were summarized in Figure 2 [7,9,23]. Briefly, descriptive statistics were obtained to determine the distribution of the screener score. Pearson’s correlation testing was performed to compare how strongly the screener score correlated with overall and component scores of Fung DASH. For components with weak correlation (r < 0.30), sensitivity analyses were conducted by using different gradations or combinations of intake questions. T-tests and analysis of variance using a Bonferroni adjustment were performed to compare screener scores of males and females, ages groups (20–39 compared with 40–59 compared with ≥60) [24], smoking status (never, former, and current), physical activity levels (sedentary, light active, and active) [25], and BMI categories (underweight, normal weight, overweight, and obese) to examine known groups validity. Weighted Cohen’s κ was used to test concordance of the screener scores across Fung DASH tertiles. All statistical procedures were conducted in Stata SE 16.0. Survey-specific procedures and dietary day 1 sample weight were used to account for unequal selection probability, clustered design, weights for nonresponse, and day of the week of recall, with type 1 error rate set at 0.05.
FIGURE 2.
Strategies used to evaluate the construct validity of the DASH Screener score. ANOVA, analysis of variance; DASH, Dietary Approaches to Stop Hypertension; Fung DASH, an established DASH adherence score; SES, socioeconomic status.
Results
Descriptive statistics of the NHANES subset of 14,651 adults are shown in Table 3. Participants were on average 47.6 (±16.9) y, 68% non-Hispanic White, 53% female, 56% nonsmokers, 31% had college degrees, and had a BMI of 28.9. 28.9% were hypertensive, and 9.4% were diagnosed with diabetes. The mean (SD) of the DASH screener score and Fung DASH score were 47.0 (14.7) and 23.8 (5.2), allowing for sufficient variation. Correlations between the DASH screener score and Fung DASH score are shown in Table 4. Correlation between the total scores was 0.70 (P < 0.001). Most components were strongly correlated (r = 0.62–1, P < 0.001), except for sodium (r = 0.04). SSB was the component that showed the strongest correlation (r = 1.00). For groups with known differences in diet quality, estimated means of DASH screener score and Fung DASH score by subpopulation are shown in Table 5. Younger adults had lower scores (44.4) than older adults (49.8); Current smokers had significantly lower screener scores (40.8) than former (48.6) or nonsmokers (48.0); people with sedentary physical activity status had lower scores (44.6) than people who were physically active (49.4). All differences were statistically significant (P < 0.001). No differences were detected between male and female (P = 0.26). For concordance, the total screener score had substantial agreement across Fung DASH score tertiles with a weighted Cohen’s κ of 0.62 (P < 0.001).
TABLE 3.
Descriptive characteristics of NHANES 2009–2014 participants (n = 14,651).
| Overall (% or mean ± SD) | |
|---|---|
| Validated DASH score (8–40) | 23.8 ± 5.2 |
| DASH screener score (0–100) | 47.0 ± 14.7 |
| Age (y) | 47.6 ± 16.9 |
| Female | 53.0 |
| Race/ethnicity | |
| Non-Hispanic White | 68.1 |
| Non-Hispanic Black | 10.9 |
| Hispanic | 13.8 |
| Education | |
| High school or less | 37.5 |
| Some college | 31.9 |
| College grad or above | 30.5 |
| Smoking status | |
| Never | 56.3 |
| Former | 24.6 |
| Current | 19.0 |
| Waist circumference, cm | 98.9 ± 14.0 |
| BMI (kg/m2) | 28.9 ± 5.9 |
| Physical activity per week1 | |
| Sedentary (MET min = 0) | 46.3 |
| Lightly active (0 < MET min < 600) | 16.5 |
| Active (MET min ≥ 600) | 37.2 |
| Total energy intake (kcal) | 2059 ± 656 |
| Hypertensive medication use | |
| Not hypertensive | 71.1 |
| No | 25.6 |
| Yes | 3.3 |
| Diabetes | |
| Yes | 9.4 |
| No | 88.3 |
| Borderline | 2.3 |
| Systolic blood pressure (mm Hg) | 121 ± 14.7 |
| Diastolic blood pressure (mm Hg) | 70.1 ± 10.4 |
| Total-cholesterol (mg/dL) | 194.0 ± 35.7 |
| Low-density lipoprotein (mg/dL) | 114.8 ± 34.8 |
| High-density lipoprotein (mg/dL) | 53.3 ± 13.8 |
| Hemoglobin A1c, % | 5.6 ± 0.8 |
Abbreviations: DASH, Dietary Approaches to Stop Hypertension; MET, metabolic equivalents of task.
Physical activity thresholds based on USDA Physical Activity Guideline for Americans, 2nd edition.
TABLE 4.
Estimated correlations (Pearson’s r) between the established DASH score and (1) the DASH screener score and (2) the DASH screener score without sodium in NHANES 2009–2014.
| Established DASH score component |
||||||||
|---|---|---|---|---|---|---|---|---|
| Fruits | Vegetables | Nuts + legumes | Dairy | Sodium | Red meat | SSB | Total | |
| DASH screener | ||||||||
| Fruits + Vegetables | 0.63 | 0.70 | 0.46 | |||||
| Legumes | 0.04 | 0.07 | 0.62 | 0.22 | ||||
| Dairy | 0.06 | 0.03 | 0.07 | 0.84 | 0.25 | |||
| Sodium | 0.01 | –0.03 | –0.01 | 0.02 | 0.04 | 0.04 | ||
| Red meat | 0.08 | –0.15 | 0.03 | 0.01 | 0.16 | 0.71 | 0.30 | |
| SSB | 0.18 | 0.08 | 0.11 | 0.02 | 0.07 | 0.11 | 1.00 | 0.55 |
| Total | 0.49 | 0.38 | 0.31 | 0.20 | –0.12 | 0.34 | 0.48 | 0.70 |
| DASH screener without sodium | ||||||||
| F+V | 0.63 | 0.70 | 0.46 | |||||
| Legumes | 0.03 | 0.07 | 0.62 | 0.22 | ||||
| Dairy | 0.06 | 0.03 | 0.07 | 0.84 | 0.25 | |||
| Red meat | 0.08 | –0.15 | 0.03 | 0.01 | — | 0.71 | 0.30 | |
| SSB | 0.18 | 0.08 | 0.11 | 0.02 | — | 0.11 | 1.00 | 0.55 |
| Total | 0.51 | 0.41 | 0.33 | 0.20 | — | 0.36 | 0.49 | 0.73 |
Abbreviations: DASH, Dietary Approaches to Stop Hypertension; SSB, sugar-sweetened beverages.
All P values < 0.0001.
TABLE 5.
Estimated mean of (1) DASH screener score, (2) DASH screener score without sodium, and (3) the established DASH score by groups with known differences in diet quality in NHANES 2009–2014.
| DASH screener1 | P value | Mean score |
P value | Fung DASH3 | P value | |
|---|---|---|---|---|---|---|
| DASH screener without sodium2 | ||||||
| Age (y) | ||||||
| 20–39 | 44.4 | Ref | 44.7 | Ref | 22.6 | Ref |
| 40–59 | 46.9 | <0.001 | 47.3 | <0.001 | 23.9 | <0.001 |
| 60+ | 49.8 | <0.001 | 49.7 | <0.001 | 25.4 | <0.001 |
| Sex | ||||||
| Male | 46.6 | Ref | 46.5 | Ref | 23.0 | Ref |
| Female | 46.9 | 0.259 | 47.4 | <0.001 | 24.6 | <0.001 |
| Race/ethnicity | ||||||
| Non-Hispanic White | 47.6 | Ref | 47.4 | Ref | 24.3 | Ref |
| Non-Hispanic Black | 41.9 | <0.001 | 42.4 | <0.001 | 21.8 | <0.001 |
| Hispanic | 45.7 | <0.001 | 47.2 | 0.591 | 23.1 | <0.001 |
| Other | 48.1 | 0.465 | 49.8 | 0.001 | 24.1 | 0.368 |
| Education | ||||||
| High school or below | 43.4 | Ref | 43.5 | Ref | 22.4 | Ref |
| Some college | 46.1 | <0.001 | 46.2 | <0.001 | 23.5 | <0.001 |
| College grad or above | 51.6 | <0.001 | 52.2 | <0.001 | 25.9 | <0.001 |
| Smoking status | ||||||
| Current smoker | 40.8 | Ref | 40.6 | Ref | 21.2 | Ref |
| Former smoker | 48.6 | <0.001 | 48.9 | <0.001 | 24.6 | <0.001 |
| Nonsmoker | 48.0 | <0.001 | 48.3 | <0.001 | 24.4 | <0.001 |
| Physical activity | ||||||
| Sedentary | 44.6 | Ref | 44.7 | Ref | 23.0 | Ref |
| Lightly active | 47.0 | <0.001 | 47.2 | <0.001 | 24.2 | <0.001 |
| Active | 49.4 | <0.001 | 49.8 | <0.001 | 24.7 | <0.001 |
| BMI category | ||||||
| Normal weight | 48.1 | Ref | 48.4 | Ref | 24.6 | Ref |
| Underweight | 45.0 | 0.038 | 44.4 | 0.010 | 23.4 | 0.014 |
| Overweight | 47.2 | 0.015 | 47.6 | 0.037 | 24.0 | <0.001 |
| Obese | 45.4 | <0.001 | 45.4 | <0.001 | 23.1 | <0.001 |
Abbreviation: DASH, Dietary Approaches to Stop Hypertension.
DASH screener score ranges from 0 to 100.
DASH Screener score without the sodium component ranges from 0 to 100.
Fung DASH, a validated DASH score, ranges from 8 to 40.
Sensitivity analysis on the DASH screener score sodium component
Sensitivity analyses were conducted to explore the lack of correlation between the sodium component in the DASH screener with the Fung DASH score. In the DASH screener, the 2 questions regarding sodium intake were “How often do you add salt to food at table” and “How often do you use salt in preparation.” Regardless of whether we used a combination of 1or 2 sodium intake questions, or whether we used different gradations (tertiles compared with quintiles), correlations between the sodium component score and mean sodium intake remained low (r = 0.03–0.05). Furthermore, we explored the mean sodium intake estimated from 24-h recall by every combination of responses to the 2 questions and compared its range with the range captured by the Fung DASH sodium score. We found that the estimated sodium intake range using a combination of responses to the 2 questions was 3297–3548 mg, which was much smaller compared with the one captured by Fung DASH sodium score (1590–5647 mg). We also examined the proportion of people grouped into each Fung sodium score quintile (Q1–Q5) for each of the 9 combinations. Among people who answered “always” to both salt use at the table and in preparation, 32% had low sodium intake (Q1 + Q2); among people who answered “rarely/never” to the questions, 40.4% had high sodium intake (Q4 + Q5), indicating that these questions do not correctly distinguish between high and low sodium consumers. See Supplemental Table 1 for detailed analysis.
On the basis of Pearson’s correlations and sensitivity analyses, it was clear that the 2 questions on sodium intake from the SAMMPRIS study did not sufficiently differentiate sodium intake. Thus, we recalculated the DASH screener score by removing the sodium component and reconducted construct validity testing. The correlation between the validated total DASH screener score and the established Fung DASH score remained strong with slight improvement (r = 0.73). The validated screener score distinguished all groups with known diet quality differences and maintained substantial concordance (κ = 0.64, P < 0.0001).
Discussion
The results provided evidence supportive of the construct validity of the brief DASH screener score. Specifically, it 1) demonstrated variation in scores across a nationally representative sample of the United States population, 2) showed strong correlation with an established DASH score for both total and component scores, 3) distinguished between groups with known differences in diet quality, and 4) showed substantial concordance with an established DASH score.
The construct validity of the DASH screener score was supported by the findings of the NHANES analyses. The wide distribution of the total DASH screener score indicated its capacity to detect meaningful differences in diet quality across the population, as well as distinguishing groups with known differences in diet quality (younger and older adults, females and men, smokers and nonsmokers, and people with different physical activity levels). In addition, the DASH screener score demonstrated strong correlation with an established DASH score, indicating that it can capture diet quality in a manner similar to a validated measure. Lastly, the high agreement between the DASH screener score and the established DASH score suggested high concordance, further demonstrating the construct validity of the DASH screener score.
Among diet screeners developed for use in clinical settings, the 14-point Mediterranean Diet Adherence Screener (MEDAS) and the Rapid Eating Assessment for Participants (REAP) are 2 frequently used and well-validated tools. The REAP questionnaire is moderately correlated with the Healthy Eating Index (r = 0.49), whereas MEDAS has been validated against FFQs (r = 0.52) and food diaries (r = 0.57) [[26], [27], [28]]. The DASH screener demonstrated a strong correlation with a validated DASH score (r = 0.73), comparable with the level reported in REAP and MEDAS.
The sodium component of the DASH screener score showed low correlation with the sodium component of an established DASH score, indicating that questions regarding salt use at the table and salt use during food preparation alone are largely insufficient to capture sodium intake. Accurate assessment of sodium intake is fraught with difficulties because of the extensive use of sodium in food processing [29]. Indeed, >70% of sodium intake in the United States came from commercially processed, packaged, and prepared foods [30], which constitutes >50% of energy intake among United States adults and continues to increase [31]. Only ∼5% of sodium intake comes from salt added at the table, and ∼6% from at-home food preparation [32]. Additionally, about one-third of food calories consumed by the United States population are from sources outside the home, which was not accounted for by asking about salt use during preparation [33]. Therefore, to obtain more accurate estimates, questions about salt intake need to be explored differently.
A few short questionnaires for dietary sodium intake assessment, with the length of the questions ranging from 43 to 26 items, have been developed and validated in South African [29], Palestinian [34], Canadian [35], and United States populations [36]. However, the predicted sodium intake from the shortest screener with 26 items was only moderately correlated with sodium intake based on 24-h recalls [37]. On the basis of these questionnaires, food-based questions, portion size estimation, along with discretionary salt use questions are needed to improve the accuracy of sodium intake assessment.
Although sodium intake is an integral component of the DASH diet, DASH diet patterns go beyond sodium intake, which encourages intake of fruit and vegetables, low-fat dairy, whole grains, and lean protein, and discourages intake of red and processed meats, foods, and beverages with added sugars, saturated fats, and sodium. Furthermore, removal of the sodium component improved correlation and concordance of the screener score with the validated DASH score and improved its ability to distinguish groups with known diet quality differences. Additional validation will be explored in future research to examine the predictive validity of the DASH screener score in both clinical and nonclinical populations.
Our study has some limitations. Like any study using self-reported data, there is measurement error associated with estimating dietary intake. However, because the DASH screener score and the established DASH score were derived from the same 24-h recall data, the risk of substantial bias is low. Our screener has limited ability to capture sodium intake, but it is a validated indicator for overall adherence to the DASH dietary pattern. Additionally, the test-retest reliability of the screener could not be examined due to limitations in the dataset.
Our study also has several strengths. We used a relatively large subset of NHANES data, which was representative of the United States population, enhancing its external validity. Our strategies for construct validity testing were informed by high-quality validation studies and guided by the AHA scientific statement on rapid dietary assessment, ensuring comparability with existing research and adherence to established validation criteria. Additionally, we performed sensitivity analyses, which allowed us to assess the robustness of our findings using alternative methods of operationalizing sodium intake, which subsequently led to the exclusion of the sodium component from the screener.
In conclusion, the brief DASH diet screener demonstrated construct validity. After removal of the sodium component, the screener remained strongly correlated and concordant with the validated DASH score and distinguished all groups with known differences in diet quality. To accurately estimate salt intake, assessment should go beyond questions about salt use at the table and during food preparation, incorporating food-based questions, portion size estimation, and discretionary salt use estimation. Future research is needed to evaluate the predictive validity of the DASH diet screener score in both clinical and nonclinical populations.
Author contributions
The authors’ responsibilities were as follows – QY, MKV: designed research; QY: performed statistical analysis, drafted the manuscript, and primary responsibility for the final content; MKV, SAC, BMO: contributed to the manuscript; and all authors: read and approved the final manuscript.
Data availability
Data described in the manuscript, code book, and analytic code will be made available on reasonable request.
Funding
QY was supported by the Dean’s Fellowship from the University of Rhode Island. MKV was supported by a K01 Career Development award from the National Heart, Lung, and Blood Institute (5K01HL165104). The funder had no role in the design, implementation, analysis, and interpretation of the data.
Conflict of interest
The authors report no conflicts of interest.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.cdnut.2025.107593.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
References
- 1.Martin S.S., Aday A.W., Almarzooq Z.I., Anderson C.A.M., Arora P., Avery C.L., et al. 2024 heart disease and stroke statistics: a report of US and global data from the American Heart Association. Circulation. 2024;149(8):e347–e913. doi: 10.1161/CIR.0000000000001209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Qiao J., Lin X., Wu Y., Huang X., Pan X., Xu J., et al. Global burden of non-communicable diseases attributable to dietary risks in 1990–2019. J. Hum. Nutr. Diet. 2022;35(1):202–213. doi: 10.1111/jhn.12904. [DOI] [PubMed] [Google Scholar]
- 3.Roth G.A., Mensah G.A., Johnson C.O., Addolorato G., Ammirati E., Baddour L.M., et al. Global burden of cardiovascular diseases and risk factors, 1990-2019: update from the GBD 2019 study. J. Am. Coll. Cardiol. 2020;76(25):2982–3021. doi: 10.1016/j.jacc.2020.11.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chiuve S.E., Fung T.T., Rexrode K.M., Spiegelman D., Manson J.A.E., Stampfer M.J., et al. Adherence to a low-risk, healthy lifestyle and risk of sudden cardiac death among women. JAMA. 2011;306(1):62–69. doi: 10.1001/jama.2011.907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sacks F.M., Svetkey L.P., Vollmer W.M., Appel L.J., Bray G.A., Harsha D., et al. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet, DASH-Sodium Collaborative Research Group. N. Engl. J. Med. 2001;344(1):3–10. doi: 10.1056/NEJM200101043440101. [DOI] [PubMed] [Google Scholar]
- 6.Shan Z., Li Y., Baden M.Y., Bhupathiraju S.N., Wang D.D., Sun Q., et al. Association between healthy eating patterns and risk of cardiovascular disease. JAMA Intern. Med. 2020;180(8):1090–1100. doi: 10.1001/jamainternmed.2020.2176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Vadiveloo M., Lichtenstein A.H., Anderson C., Aspry K., Foraker R., Griggs S., et al. Rapid diet assessment screening tools for cardiovascular disease risk reduction across healthcare settings: a scientific statement from the American Heart Association. Circ. Cardiovasc. Qual. Outcomes. 2020;13(9) doi: 10.1161/HCQ.0000000000000094. [DOI] [PubMed] [Google Scholar]
- 8.Kirkpatrick S.I., Reedy J., Butler E.N., Dodd K.W., Subar A.F., Thompson F.E., et al. Dietary assessment in food environment research: a systematic review. Am. J. Prev. Med. 2014;46(1):94–102. doi: 10.1016/j.amepre.2013.08.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Guenther P.M., Kirkpatrick S.I., Reedy J., Krebs-Smith S.M., Buckman D.W., Dodd K.W., et al. The healthy eating index-2010 is a valid and reliable measure of diet quality according to the 2010 dietary guidelines for Americans. J. Nutr. 2014;144(3):399–407. doi: 10.3945/jn.113.183079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Turan T.N., Lynn M.J., Nizam A., Lane B., Egan B.M., Le N.A., et al. Rationale, design, and implementation of aggressive risk factor management in the stenting and aggressive medical management for prevention of recurrent stroke in intracranial stenosis (SAMMPRIS) trial. Circ. Cardiovasc. Qual. Outcomes. 2012;5(5):e51–e60. doi: 10.1161/CIRCOUTCOMES.112.966911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Gordon N.F., Salmon R.D., Franklin B.A., Sperling L.S., Hall L., Leighton R.F., et al. Effectiveness of therapeutic lifestyle changes in patients with hypertension, hyperlipidemia, and/or hyperglycemia. Am. J. Cardiol. 2004;94(12):1558–1561. doi: 10.1016/j.amjcard.2004.08.039. [DOI] [PubMed] [Google Scholar]
- 12.Fung T.T., Chiuve S.E., McCullough M.L., Rexrode K.M., Logroscino G., Hu F.B. Adherence to a DASH-style diet and risk of coronary heart disease and stroke in women. Arch. Intern. Med. 2008;168(7):713–720. doi: 10.1001/archinte.168.7.713. [DOI] [PubMed] [Google Scholar]
- 13.CDC About the National Health and Nutrition Examination Survey. https://www.cdc.gov/nchs/nhanes/about/?CDC_AAref_Val=https://www.cdc.gov/nchs/nhanes/about_nhanes.htm [Internet]. [cited 2024 Nov 5]. Available from:
- 14.Raper N., Perloff B., Ingwersen L., Steinfeldt L., Anand J. An overview of USDA’s dietary intake data system. J. Food Compos. Anal. 2004;17:545–555. [Google Scholar]
- 15.CDC National Health and Nutrition Examination Survey 2009-2010 Data Documentation, Codebook, and Frequencies Dietary Interview—Individual Foods. First Day [Internet] 2012 https://wwwn.cdc.gov/Nchs/Nhanes/2009-2010/DR1IFF_F.htm [cited 2024 Nov 5]. Available from: [Google Scholar]
- 16.Sotos-Prieto M., Bhupathiraju S.N., Mattei J., Fung T.T., Li Y., Pan A., et al. Changes in diet quality scores and risk of cardiovascular disease among US men and women. Circulation. 2015;132(23):2212–2219. doi: 10.1161/CIRCULATIONAHA.115.017158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Banna J.C., McCrory M.A., Fialkowski M.K., Boushey C. Examining plausibility of self-reported energy intake data: considerations for method selection. Front. Nutr. 2017;4:45. doi: 10.3389/fnut.2017.00045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Food Surveys Research Group Food Patterns Equivalents Database Overview. https://www.ars.usda.gov/northeast-area/beltsville-md-bhnrc/beltsville-human-nutrition-research-center/food-surveys-research-group/docs/fped-overview/ [Internet]. [cited 2024 Nov 5]. Available from:
- 19.Reedy J., Lerman J.L., Krebs-Smith S.M., Kirkpatrick S.I., Pannucci T.R.E., Wilson M.M., et al. Evaluation of the Healthy Eating Index-2015. J. Acad. Nutr. Diet. 2018;118(9):1622–1633. doi: 10.1016/j.jand.2018.05.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zhao Y., Araki T. Diet quality and its associated factors among adults with overweight and obesity: findings from the 2015-2018 National Health and Nutrition Examination Survey. Br. J. Nutr. 2024;131(1):134–142. doi: 10.1017/S0007114523001587. [DOI] [PubMed] [Google Scholar]
- 21.Armstrong T., Bull F. Development of the World Health Organization Global Physical Activity Questionnaire (GPAQ) J. Public Health. 2006;14:66–70. [Google Scholar]
- 22.Piercy K.L., Troiano R.P., Ballard R.M., Carlson S.A., Fulton J.E., Galuska D.A., et al. The physical activity guidelines for Americans. JAMA. 2018;320(19):2020–2028. doi: 10.1001/jama.2018.14854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Parker H.W., De Araujo C., Thorndike A.N., Vadiveloo M.K. The utility of household Grocery Purchase Quality Index scores as an individual diet quality metric. Br. J. Nutr. 2021;126(6):933–941. doi: 10.1017/S0007114520004833. [DOI] [PubMed] [Google Scholar]
- 24.Mellen P.B., Gao S.K., Vitolins M.Z., Goff D.C. Deteriorating dietary habits among adults with hypertension: DASH dietary accordance, NHANES 1988-1994 and 1999-2004. Arch. Intern. Med. 2008;168(3):308–314. doi: 10.1001/archinternmed.2007.119. [DOI] [PubMed] [Google Scholar]
- 25.Leskinen T., Stenholm S., Pulakka A., Pentti J., Kivimäki M., Vahtera J. Comparison between recent and long-term physical activity levels as predictors of cardiometabolic risk: a cohort study. BMJ Open. 2020;10(2) doi: 10.1136/bmjopen-2019-033797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Schröder H., Fitó M., Estruch R., Martínez-González M.A., Corella D., Salas-Salvadó J., et al. A short screener is valid for assessing Mediterranean diet adherence among older Spanish men and women. J. Nutr. 2011;141(6):1140–1145. doi: 10.3945/jn.110.135566. [DOI] [PubMed] [Google Scholar]
- 27.García-Conesa M.T., Philippou E., Pafilas C., Massaro M., Quarta S., Andrade V., et al. Exploring the validity of the 14-item Mediterranean diet adherence screener (Medas): a cross-national study in seven European countries around the Mediterranean region. Nutrients. 2020;12(10):2960. doi: 10.3390/nu12102960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Gans K.M., Risica P.M., Wylie-Rosett J., Ross E.M., Strolla L.O., McMurray J., et al. Development and evaluation of the nutrition component of the Rapid Eating and Activity Assessment for Patients (REAP): a new tool for primary care providers. J. Nutr. Educ. Behav. 2006;38(5):286–292. doi: 10.1016/j.jneb.2005.12.002. [DOI] [PubMed] [Google Scholar]
- 29.Charlton K.E., Steyn K., Levitt N.S., Jonathan D., Zulu J.V., Nel J.H. Development and validation of a short questionnaire to assess sodium intake. Public Health Nutr. 2008;11(1):83–94. doi: 10.1017/S1368980007000146. [DOI] [PubMed] [Google Scholar]
- 30.Quader Z.S., Zhao L., Gillespie C., Cogswell M.E., Terry A.L., Moshfegh A., et al. Sodium intake among persons aged ≥2 years—United States, 2013–2014. MMWR Morb. Mortal. Wkly. Rep. 2017;66(12):324–328. doi: 10.15585/mmwr.mm6612a3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Juul F., Parekh N., Martinez-Steele E., Monteiro C.A., Chang V.W. Ultra-processed food consumption among US adults from 2001 to 2018. Am. J. Clin. Nutr. 2022;115(1):211–221. doi: 10.1093/ajcn/nqab305. [DOI] [PubMed] [Google Scholar]
- 32.Bibbins-Domingo K., Chertow G.M., Coxson P.G., Moran A., Lightwood J.M., Pletcher M.J., et al. Projected effect of dietary salt reductions on future cardiovascular disease. N. Engl. J. Med. 2010;362(7):590–599. doi: 10.1056/NEJMoa0907355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Lin Biing-Hwan, Guthrie Joanne, Smith Travis. Dietary Quality by Food Source and Demographics in the United States, 1977–2018. EIB-249. U.S. Department of Agriculture, Economic Research Service; March 2023. [Google Scholar]
- 34.Awwad R., Qasrawi R., Alwafa R.A., Kharaz L., Badrasawi M. Development and validation of food frequency questionnaire screener software for sodium intake among Palestinian population. BMC Nutr. 2024;10(1):120. doi: 10.1186/s40795-024-00927-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Cooper M., Simpson J.R., Klutka R. Development and validation of a sodium AnaLysis tool (SALT) Nutr. J. 2020;19(1):55. doi: 10.1186/s12937-020-00555-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Block G., Hartman A.M., Dresser C.M., Carroll M.D., Gannon J., Gardner L. A data-based approach to diet questionnaire design and testing. Am. J. Epidemiol. 1986;124(3):453–469. doi: 10.1093/oxfordjournals.aje.a114416. [DOI] [PubMed] [Google Scholar]
- 37.Tangney C.C., Rasmussen H.E., Richards C., Li M., Appelhans B.M. Evaluation of a brief sodium screener in two samples. Nutrients. 2019;11(1):166. doi: 10.3390/nu11010166. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data described in the manuscript, code book, and analytic code will be made available on reasonable request.