The Dutch Healthy Diet Index : development, evaluation, and application

The Dutch Healthy Diet index – Development, Evaluation, and Application

Linde van Lee

Abstract

Background: Dietary indices evaluate the conformity of an individual’s diet with pre-defined standards. Generally, dietary guidelines are used for this purpose. As no index based on the current dietary guidelines was available in the Netherlands, the aim of the present thesis was to develop, evaluate, and apply a dietary index for use in the country.

Methods and results: The Dutch Healthy Diet index (DHD-index) was developed on the basis of the 2006 Dutch dietary guidelines using data relating to 749 young adults who completed two 24-hour recalls in the Dutch national food consumption survey 2003. The index comprises ten components on physical activity, vegetables, fruit, dietary fibre, saturated fatty acids, trans fatty acids, consumption occasions with acidic drinks and foods, sodium, and alcohol. Scores for each component range between 0 (no adherence) and 10 (complete adherence) points. The DHD-index was inversely associated with energy intake and positively associated with most micronutrient intakes when adjusted for energy intake. We compared the DHD-index score based on two 24-hour recalls with the index based on the food frequency questionnaires (FFQ) of 121 adults from the European Food Consumption Validation study. We revealed an acceptable correlation (r=0.48) and absolute agreement between the indices based on the two methods. The prospective relationship with mortality outcomes was studied in 3593 of the Rotterdam Study participants who were followed for 20 years. The DHD-index per 10 points increment was associated with a 9% (95% CI 0.87-0.96) risk reduction for all-cause mortality, and non-significantly associated with risk reductions for cardiovascular disease, coronary heart disease, and stroke mortality. Among women, shared dinners were associated with lower DHD-index scores for that day than solo dinners in 1740 participants who contributed multiple 24-hour recalls in the Nutrition Questionnaires plus study. Among men and women, dinners shared with family members were associated with a higher DHD-index score on that day than dinners shared with others. Furthermore, in a subsample of 1235 participants in the Nutrition Questionnaires plus study, we evaluated the DHD-index based on the newly developed 34-item DHD-FFQ, a short questionnaire to assess diet quality in time-limited settings. The DHD-index based on the DHD-FFQ showed an acceptable correlation (r=0.56) with the index based on a 180-item FFQ, but showed a large variation in bias at individual level.

Conclusions: The DHD-index based on an FFQ, on multiple 24-hour recalls, or on the DHD-FFQ was considered a valid tool to rank participants according to their diet quality. The DHD-index was therefore considered useful to monitor populations, study diet–disease associations, and identify subpopulations at risk of poor diet quality.