Higher dietary diversity scores do not guarantee healthier or more sustainable diets

A higher dietary diversity score is often assumed to signal a healthier, more nutritionally adequate, or more sustainable diet. The review, published in Frontiers in Sustainable Food Systems on 29 October 2026, traces how dietary diversity indicators have evolved over nearly five decades and concludes that the scientific value of any such indicator lies not in its complexity but in the alignment between what it actually measures and the outcome it is intended to represent.

The review was conducted by Brian Kipkoech of the Free University of Bozen-Bolzano, Tensae Berhanu Fororo, also of Bozen-Bolzano, and Mercy Kilel of the French National Institute for Agriculture, Food and Environment. The authors searched Scopus, Web of Science, and PubMed for foundational and contemporary publications on the conceptual development, measurement, validation, and application of dietary diversity indicators, supplemented by reference screening and institutional guidance from organisations such as the Food and Agriculture Organisation and the World Health Organisation. In total, 28 literature sources published between 1978 and 2026 informed the synthesis. Because the aim was to analyse conceptual and methodological progress rather than to calculate pooled associations between diversity and specific health outcomes, the authors chose a narrative-conceptual review design instead of a systematic review with meta-analysis.

Dietary diversity is defined as the variety of distinct foods or food groups consumed within a specified reference period. Its nutritional rationale is straightforward: no single food supplies all the macro- and micronutrients required for optimal health, so eating a wider range of foods increases the likelihood of meeting nutrient needs. This logic has made diversity indicators attractive as practical, low-cost proxies for assessing dietary adequacy, monitoring food security, and running public health programmes. Yet, as the review emphasises, interpretation depends critically on the indicator employed, the population assessed, the unit of analysis, and the nutritional or socioeconomic outcome of interest.

From the literature, the authors distinguished six overlapping methodological categories. Food-count-based indicators, the oldest approach, operationalise diversity as the number of distinct foods or food groups consumed in a reference period, typically from 24-hour dietary recalls. Composite and dietary guideline-based indicators add dimensions such as evenness, dominance, dissimilarity, and adherence to dietary recommendations. Nutritional indicators link diversity directly to nutrient adequacy. Sustainability- and biodiversity-oriented indicators, such as Dietary Species Richness, count the distinct biological species consumed. Temporal indicators capture how variety fluctuates across days, seasons, and longer periods. Finally, personalised and biological approaches, drawing on precision nutrition, relate diversity to individual genetics, metabolism, and gut microbiota.

Each category measures a distinct construct at a different analytical scale. Food Variety Scores quantify the number of distinct foods, whereas food-group-based indicators characterise the distribution of consumption across predefined categories. Household Dietary Diversity Scores are primarily intended to reflect household food access, while indicators such as Minimum Dietary Diversity for Women serve as population-level proxies for micronutrient adequacy. Ecological indices adapted to diets capture two further components: richness, the total number of unique food items, and evenness, how evenly food proportions are distributed. These measures are related but not interchangeable, since their values depend on different classification systems and levels of food aggregation.

The review highlights why identical scores can conceal radically different diets. Food-count indicators weight all foods or food groups equally, regardless of portion size, nutrient density, processing level, or proportional representation. A diet can therefore reach the same diversity score as another while deriving much of its variety from energy-dense, nutrient-poor, or ultra-processed foods. Greater variety can improve nutrient adequacy, but it may also stem from exactly the dietary transitions—rising consumption of ultra-processed, energy-dense foods—that fuel diet-related non-communicable diseases. Diets with comparable diversity scores may consequently differ substantially in nutritional quality, energy density, dietary balance, and associations with metabolic health.

Composite indices attempt to address these weaknesses. The Shannon entropy index captures both richness and evenness, while Simpson-based measures place greater emphasis on the dominance of frequently consumed components, allowing diets dominated by a few foods to be distinguished from more evenly distributed ones even when the total food count is identical. The Healthy Food Diversity Index goes further by integrating food variety, the proportional distribution of consumed foods, and their health value, and analyses showed that incorporating health-related weights improved correspondence with dietary quality compared with selected conventional measures. Yet these approaches introduce their own problems: they are sensitive to how food categories are defined and to the consumption metric used, whether dietary energy, weight, or frequency, and incorporating nutritional weights introduces normative assumptions that can blur the line between diversity and overall diet quality. Dietary guidelines also differ across populations and change over time, complicating longitudinal and cross-country comparisons.

Nutritional indicators have strengthened the biological interpretation of diversity. Validation studies established associations between food-group diversity and micronutrient adequacy, including evidence that simple food-group indicators predicted micronutrient adequacy among women across five resource-poor settings and supported a population-level threshold for women of reproductive age. A systematic scoping review cited in the article found more consistent positive associations between dietary diversity indicators and micronutrient adequacy than with broader health outcomes, underscoring the need to distinguish nutritional validation from clinical relevance. On the biodiversity front, Dietary Species Richness measures the number of distinct biological species consumed and has demonstrated relevance to micronutrient adequacy, capturing traditional and underutilised species that conventional food-group classifications overlook. The authors caution, however, that dietary biodiversity should not be read as direct evidence of nutritional adequacy or environmental sustainability; those outcomes require complementary indicators and independent validation.

Temporal and personalised approaches extend the picture further. Because dietary variety fluctuates across days and seasons, single 24-hour recalls cannot characterise habitual patterns; repeated assessments can distinguish people who consistently eat diverse diets from those who achieve similar cumulative variety through occasional consumption. But diversity estimates depend on the observation period, assessment frequency, and treatment of repeated foods, and dietary repetition is not inherently undesirable, since regularly consumed nutrient-dense foods can support adequate diets. In precision nutrition, molecular techniques such as DNA metabarcoding can identify consumed biological taxa from stool samples, offering complementary information on dietary composition. No standardised, validated precision dietary diversity indicator yet exists, however, and molecular detection does not establish consumption quantities or nutrient adequacy, while the data demands of personalised assessment limit its use in large-scale surveillance.

The review’s central message is a call for conceptual and methodological precision. Food richness, food-group representation, proportional evenness, nutritional dissimilarity, and dietary guideline adherence are distinct constructs that can produce divergent assessments of the same diet, and inconsistencies in food classification, reference periods, consumption thresholds, weighting procedures, and analytical scales compound the problem. Household food access, individual consumption, and national food availability are fundamentally different levels of observation, and conclusions at one level cannot be transferred to another. The authors recommend that policymakers select indicators according to clearly defined objectives, that monitoring frameworks harmonise indicator definitions, food-group classifications, reference periods, and reporting standards, and that diversification policies distinguish between increasing the number of foods available and improving the nutritional composition, accessibility, affordability, and sustainability of diets. Dietary diversity, they conclude, should be treated as a purpose-specific component of nutritional and food-system assessment rather than a universal substitute for direct measures of dietary adequacy, health, or sustainability.

Subject of Research: Evolution, measurement, and interpretation of dietary diversity indicators in nutrition and food-system research

Article Title: Dietary diversity indicators: evolution, measurement, and interpretative challenges

Article References: Kipkoech, B., Berhanu Fororo, T., & Kilel, M. (2026). Dietary diversity indicators: evolution, measurement, and interpretative challenges. Frontiers in Sustainable Food Systems, 10, 1954641. https://doi.org/10.3389/fsufs.2026.1954641

Image Credits: AI Generated

DOI: 10.3389/fsufs.2026.1954641

Keywords: dietary diversity, nutrition, food security, nutrient adequacy, diet quality, biodiversity, sustainable diets, precision nutrition, dietary assessment, food systems, public health, measurement validity