Validity of a short method based on food frequency and multiple regression to evaluate the nutrient intakes of French Canadian Women
1986
Richard, L. | Roberge, A.G.
A short method of dietary analysis, based on food group frequencies and multiple regression, was developed and validated from 3-day dietary records from a group of 306 French Canadian divided in two independent groups of 173 (1a) and 133 (2a) women respectively. From each group, subsamples were formed to calculate and applied regression equations. The regression equations were calculated from the first group data of 173 women (1a) and these subsamples data of 87 (1b) and 45 (1c) women. These equations were calculated for energy, protein, fat, complex carbohydrates, concentrated sugars, alcohol, calcium, iron, retinol, thiamine, riboflavin, niacin and ascorbic acid. The coefficient of determination (R2) ranged from 0.15 to 0.95 and regession equations were applied to a second group of 133 (2a) and a subsample, from this group, of 87 (2b) women, respectively. The correlations between the mean energy and nutrient intakes obtained by the standard and by the short method varied from 0.04 to 0.94. The present findings pointed out that the short method is valid to study the nutrient intakes of a group of subjects because the mean results were similar to those obtained from a 3-day food record.
显示更多 [+] 显示较少 [-]