An official website of the United States government
The Healthy Eating Index (HEI) has been evaluated by assessing its psychometric properties, including content validity, construct validity, and reliability. Since 2005, analyses have been conducted to evaluate the HEI’s psychometric performance.
To do this, we scored dietary data from the most recent National Health and Nutrition Examination Survey (NHANES) for each version of the HEI. Data from 2001-2002 NHANES were used to evaluate HEI-2005; data from NHANES 2003-2004 for HEI-2010; data from NHANES 2011-2012 for HEI-2015; data from NHANES 2017-2018 for HEI-2020; data from NHANES 2011-2018 were pooled to create sufficient sample size for evaluating HEI-Toddlers-2020.
We also scored several sets of exemplary sample menus. For the HEI for 2 years and older, this has included the United States Department of Agriculture's (USDA) sample menus, the National Heart, Lung, and Blood Institute's (NHLBI) DASH program, Harvard Medical School's Healthy Eating Pyramid
, and the American Heart Association's No-Fad Diet. For the HEI-Toddlers-2020 (for 12 through 23 months), the exemplary menus were from the American Academy of Pediatrics and Healthy Eating Research.
For complete details, please refer to the following publications:
- Lerman JL, Herrick KA, Pannucci TE, Shams-White MM, Kahle LL, Zimmer M, Meyers Mathieu K, Stoody E, Reedy J. Evaluation of the Healthy Eating Index-Toddlers-2020. J Nutr Acad Diet. 2023 May 15;S2212-2672(23)00245-9. Online ahead of print.
- Reedy J, Lerman JL, Krebs-Smith SM, Kirkpatrick SI, Pannucci TE, Wilson MM, Subar AF, Kahle LL, and Tooze JA. Evaluation of the Healthy Eating Index-2015. J Acad Nutr Diet. 2018 Sep;118(9):1622-1633.
- Guenther PM, Kirkpatrick SI, Reedy J, Krebs-Smith SM, Buckman DW, Dodd KW, Casavale KO, Carroll RJ. The Healthy Eating Index-2010 is a valid measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr. 2014 Mar;144 (3):399-407.
- Guenther PM, Reedy J, Krebs-Smith SM, Reeve BB. Evaluation of the Healthy Eating Index–2005. J Am Diet Assoc. 2008 Nov;108(11):1854-64.
Evaluation of the HEI for 2 years and older
The table below describes how the HEI for 2 years and older has been evaluated. Overall, the results demonstrated evidence supportive of validity and reliability.
| Question | Strategy | Expectation | Results |
|---|---|---|---|
| Content Validity1 | |||
| 1. Do the HEI components reflect key food choice recommendations of the DGA? | Checked against the associated DGA. | HEI components will reflect the key recommendations of the DGA. | For HEI-2005 through HEI-2020, all the key food choice recommendations of the Dietary Guidelines that relate to diet quality are reflected in HEI components. By design, the HEI does not cover other Dietary Guidelines recommendations, such as those for physical activity, body weight, and food safety. |
| Construct Validity | |||
| Does the index yield high scores for exemplary menus? | Computed scores of sample menus from the United States Department of Agriculture (USDA), the National Heart, Lung, and Blood Institute's (NHLBI) DASH program, Harvard Medical School's Healthy Eating Pyramid, and the American Heart Association's No-Fad Diet. |
Exemplary menus will receive high quality scores. | Menus scored between 87.8 and 100 using HEI-2105. Menus that exclude a food group, such as Dairy, scored lower than other menus. As the HEI was refined over time, some menus scored lower on later versions of the HEI, reflecting the refinement of advice on subgroups such as dark green vegetables and components such as added sugar. |
| Does the index allow for sufficient variation in scores among individuals? | Estimated percentiles of HEI total score and component scores using NHANES data. | A full range of scores will be seen across the distribution for the total score and most dietary components. | A new addition to the evaluation of construct validity in 2010 came with a new statistical method known as the Markov Chain Monte Carlo (MCMC) method, which allowed us to estimate distributions of HEI scores of estimated usual intake in the population (rather than only the mean scores) to assess whether the distribution was wide enough to detect meaningful differences. In all versions of the index, there was sufficient variation in scores across individuals of a population. |
| Does the index differentiate between groups with known differences in diet quality? | Conducted t-tests to compare groups, such as men vs. women, older vs. younger adults, smokers vs. non-smokers. | Example: Smokers will have lower HEI scores than non-smokers. | Example: One-day scores between smokers and non-smokers were significantly different in all versions, indicating that the HEI can distinguish between groups with known differences in the quality of their diets. |
| Does the index assess diet quality independent of diet quantity? | Estimated Pearson correlations between component scores and energy intake. | Low correlation between scores and energy intake. | Low correlations were found between total and component scores and energy, indicating that the HEI can assess diet quality independently of diet quantity. |
| Is the index multidimensional? | Conducted a principal components analysis using NHANES data. | No single factor explains the total score. | Principal components analysis (PCA) demonstrated that multiple factors underlie each version of the HEI such that both the individual components and the total score provide insights into diet quality. Therefore, no single dietary component explains the total HEI score. |
| Concurrent Validity | |||
| How does an updated HEI compare to previous version(s) of the HEI? | Calculated and compared component and total scores with different HEIs for the same population. | The scores calculated will be similar but not the same. Component scores will reflect differences where guidance is notably different or has changed significantly. | Depending on the HEI, there are differences in component scores and total scores. |
| Criterion Validity | |||
| Can the index predict a health outcome? | Estimated hazard of mortality outcomes using data from the NIH-AARP Diet and Health Study. | Diet quality scores from the HEI are associated with mortality from diet-related disease. | The HEI has demonstrated predictive validity with mortality among older US men and women. For the HEI-2015, this was in the range of 13% to 23% lower risk of mortality associated with high diet quality. |
| Reliability | |||
| How internally consistent is the score? | Used Cronbach's coefficient alpha to assess internal consistency, which is the degree to which multiple components within an index measure the same underlying construct. | Because the score is multidimensional, a moderate value of internal consistency is expected because Cronbach’s alpha assumes unidimensionality. | Cronbach’s alpha was moderate; for HEI-2015 Cronbach’s alpha=0.67. The moderate results suggest that individual components provide additional insights into the quality of the diet beyond what can be interpreted from the total score alone. |
1 The HEI-2020 was evaluated for content validity only as there were no changes to the dietary components or scoring standards between the HEI-2015 and HEI-2020.
Evaluation of the HEI-Toddlers (for 12 through 23 months)
Complete details on the evaluation of the HEI-Toddlers-2020 are reported in the following publication:
- Lerman JL, Shams-White MM, Kahle L, Pannucci T, Herrick KA, Zimmer M, Meyers Mathieu K, Stoody E, Reedy J. Evaluation of the Healthy Eating Index-Toddlers-2020. J Nutr Acad Diet. 2023 May 15;S2212-2672(23)00245-9. Online ahead of print.
In brief, the HEI for toddlers 12 through 23 months was evaluated as shown below. Overall, results demonstrated evidence supportive of validity and reliability.
| Question | Strategy | Expectation |
|---|---|---|
| Construct Validity | ||
| Does the index yield high scores for exemplary menus? And conversely, does the index yield low scores for poor diet quality intake days? |
Calculated the HEI total and component scores of sample menus from the American Academy of Pediatrics and Healthy Eating Research. Reviewed foods reported on 24-hour recalls in NHANES1 2011-2018 for a random sample of days that scored less than 40 total points. | Exemplary menus receive high quality scores. Conversely, low index scores will be characterized by poor diet-quality food intake. |
| Does the index allow for sufficient variation in scores among individuals? | Estimated percentiles of component and total scores in NHANES 2011-2018. | Full range of scores will be seen across the distribution for most components and for the total score. |
| Does the index assess diet quality independent of diet quantity? | Estimated Pearson correlations between component scores and energy intake in NHANES 2011-2018. | Low correlation between scores and energy intake. |
| Is the index multidimensional? | Conducted a principal components analysis in NHANES 2011-2018. | No single factor explains the total score. |
| Concurrent Validity | ||
| How does the HEI-Toddlers-2020 compare to the HEI-2020? | Calculated HEI-Toddlers-2020 score and HEI-2020 score for the same 24-month-old subset of the population. | The total score calculated with these two different indices will be similar but not the same. Component scores will reflect differences where guidance is notably different for toddlers vs. 2 years and older, such as Added Sugars and Saturated Fat. |
| Reliability | ||
| What are the relationships among the index components and total score? What components exert the most influence on the total score? | Estimated Pearson correlations between component scores, and item-rest correlations in NHANES 2011-2018. | Reflects distinct dietary components, with low correlations between unrelated components and moderate to high correlations between coupled and related components. |
| How internally consistent is the score? | Calculated Cronbach’s alpha in NHANES 2011-2018. | Because the score is multidimensional, a moderate value of internal consistency is expected because Cronbach’s alpha assumes unidimensionality. |
1NHANES, National Health and Nutrition Examination Survey