Agreement between two Meal Quality Indexes adapted to evaluate food consumption of institutionalized elderly

Abstract

Objective
To assess agreement between two meal quality indexes adapted for elderly people, evidencing their similarities and differences.

Methods
Agreement study with 322 institutionalized elderly individuals. Food consumption data were collected by the method of weighted food record, in two nonconsecutive days. The Meal Quality Index and the Main Meal Quality Index of each older adult were calculated for the lunch meal. Agreement between methods was tested by cross-classification in quartiles and weighted kappa (Kw), and the difference between medians by the Wilcoxon test.

Results
The Meal Quality Index median was 54.67 points and the Main Meal Quality Index 53.51 points (p=0.723). When the components of each index were assessed, those associated to the consumption of carbohydrates, total fat and saturated fat were similar. The consumption of fruits separated from vegetables in the Main Meal Quality Index evidenced low consumption of vegetables by the elderly. Cross-classification by quartiles showed  good agreement; the exact one being 48.8% and the disagreement 3.4% (Kw=0,447). The proportion of elderly individuals in the same exact or adjacent quartile was greater than 85.0%. Agreement was higher in males (89.4%, Kw=0.475); in the age range of 70 to 79 years (91.1%; Kw=0.562) and in non-profit nursing homes
(96.7%; Kw=0.62).

Conclusion
The two indexes reviewed show a good agreement between them and common characteristics. The number of components is higher in the MMQI and may represent a more detailed assessment of meal quality.