7 Dietary Assessment and Nutrition Evidence
Learn how dietary assessment methods estimate food and nutrient intake, how to interpret their limitations, and how to evaluate nutrition evidence.
Choosing a dietary assessment method
Dietary assessment estimates what people eat and drink, how much they consume, and how intake relates to health. No single method measures diet perfectly, so the appropriate method depends on the question: whether it concerns a recent meal, a person's usual diet, a population's average intake, or a relationship between diet and health.
Common dietary assessment methods
A uses a trained interviewer to ask a person to report all foods and drinks consumed during the previous day. Standardized prompts and portion-size aids help capture ingredients, amounts, and preparation details. One recall describes one day, not necessarily a person's . Repeated recalls on nonconsecutive days better represent and can capture differences between weekdays, weekends, and seasons.
A asks a person to record foods and amounts as they are consumed, often for several days. Records can provide detailed information, but they require effort and may change food choices because intake is being recorded. Amounts may be estimated or weighed.
A asks how often listed foods and drinks were consumed over a longer period, such as the past month or year. It can help rank people by longer-term eating patterns in large studies, but it depends on memory and a preset food list and can have substantial systematic error.
A diet history or brief screener asks about customary eating or selected foods and nutrients. Screeners are efficient for focused questions, but generally do not measure the entire diet or total energy intake.
are measurements in blood, urine, or other tissues that can provide objective evidence about exposure to some nutrients or foods. A recovery biomarker can estimate intake for certain dietary components, but are not available or equally informative for every component.
Large surveys may combine interviewer-administered recalls with standardized food-composition databases. Combining methods can balance detail, coverage, cost, and burden.
Interpreting intake and reference values
Before interpreting dietary data, clarify what they represent: reported intake on a particular day, an estimate of , a nutrient amount, a food group, a dietary pattern, or an association with a health outcome. Converting foods into nutrient estimates depends on portion sizes, recipes, food-preparation details, and the accuracy of the food-composition database. Include supplements when the question concerns total nutrient intake; food-only estimates do not capture them.
A single day's intake is weak evidence about an individual's because people naturally eat differently from day to day. For a group, repeated measurements and statistical methods can help separate within-person variation from differences in . Group averages can conceal people with very high or very low intakes, so examine distributions as well as means when possible.
Nutrient reference values serve different purposes. The is used to assess the prevalence of inadequate intake in a group when usual-intake distributions are appropriately estimated. The is designed to meet the needs of nearly all healthy individuals in a life-stage group; it is not a suitable cut point for estimating a group's prevalence of inadequacy.
The is used when evidence is insufficient to establish an EAR and RDA. Intake below the AI does not by itself establish inadequacy. The is a safety reference, not a target. These reference values depend on characteristics such as age and life stage, and dietary data are only one part of nutritional assessment.
For example, if a person reports less than the RDA for a nutrient on one day, that alone does not show that their is inadequate. Consider repeated intake data, supplements, relevant reference values, and other clinical information.
Sources of measurement error
Dietary measurements can contain , which adds unpredictable variation, and , which shifts estimates consistently in a direction. Common sources include:
Forgetting or omitting foods, drinks, snacks, or additions such as oils, sauces, and sugar.
Misremembering how often foods were eaten or estimating portion sizes inaccurately.
Underreporting or overreporting because of social desirability, burden, or difficulty recalling intake.
Changing food choices because intake is being recorded.
Sampling too few days to represent day-to-day, weekday–weekend, or seasonal variation.
Incomplete food lists, inaccurate recipe or brand matches, and errors in nutrient databases.
Applying a questionnaire or database to a population whose foods and eating patterns differ from those for which it was developed.
Repeated short-term measurements can reduce the influence of day-to-day random variation, but repetition alone does not remove systematic bias. Validation studies compare an instrument with an appropriate reference measure. Recovery can help assess or correct error for some nutrients. No self-report tool should automatically be treated as a perfect measure of true intake.
Evaluating nutrition evidence
To evaluate a nutrition claim, define the population, dietary exposure or intervention, comparison, and health outcome. Then assess the evidence in context.
Study design: Randomized trials can test assigned interventions, while cohort studies can track dietary patterns and later outcomes. Observational associations may reflect or . A single cross-sectional snapshot generally cannot establish which came first.
Measurement quality: Ask how diet was assessed, whether the method fits the research question, and how measurement error was handled. Error can obscure real relationships or, in some cases, distort them.
Totality and consistency: Give more weight to transparent systematic reviews that search, assess, and synthesize relevant studies than to an isolated result. Consider whether findings are consistent across study designs and populations, and whether limitations or competing explanations remain.
Meaning and applicability: Distinguish relative from absolute effects, assess the size and uncertainty of an effect, and ask whether the study population, dietary pattern, and outcome apply to the question at hand.
Nutrition is often studied as a pattern of foods eaten together, not only as isolated nutrients. Foods and nutrients can interact, and replacing one food with another changes how an association should be interpreted. Systematic reviews make choices and the strength of conclusions more transparent, but they do not eliminate limitations in the underlying studies.