Objectives: To design and validate a questionnaire to identify stroke patients with an unhealthy diet.
Hypothesis: A simple questionnaire can accurately identify patients with dietary patterns associated with increased risk for vascular disease.
Methods: An expert panel identified the components of a healthy diet thought to be important for secondary stroke prevention and drafted ten candidate survey questions. A focus group of four stroke patients (2 men, 2 women) from Yale New Haven Hospital reviewed the questions to refine the wording. From this group, the investigator chose five questions to create the Fast Assessment of Diet in Stroke (FADS) score. To validate the score, among 20 patients with acute ischemic stroke, we compared FADS score to the score on an established “gold standard” measure of adherence to the Mediterranean-type diet, the Mediterranean Diet Score (MDS). The MDS has a range of 0 (lowest adherence to the diet) to 9 (highest adherence). We used ROC curve analysis to identify the FADS score with optimal performance. Based on population means in non-Mediterranean countries, we defined an unhealthy diet as an MDS score <4.
Results and Conclusions: The mean MDS score in the population was 3.2 (range 2 to 5). 74% of participants were defined as having an unhealthy diet (MDS <4). The aROC was 0.77. A FADS score ≤3 had had the highest sum for sensitivity and specificity (sensitivity = 79%, specificity = 50%). However, for clinical purposes, a FADS score of <3 may be most effective in selecting patients who need dietary counseling (sensitivity = 46%, specificity = 100%). The FADS score takes less than 2 minutes to administer and accurately identifies stroke patients who need dietary counseling. The low sensitivity, however, indicates that further development is required and that testing against other established gold standards of dietary quality should be completed. This work is under way.