Comparison of nutrient intake data calculated on the basis of two different databases. Results and experiences from a...

In international surveys of food consumption and nutrient intake, it is essential that the dietary data are comparable when different databases and calculation programs are used. The aim of the present analysis was to examine the comparability of nutrient intake data calculated on the basis of the Swedish food composition database PC-kost and the Finnish food composition database Nutrica.

A total of 20 male adults currently living in Sweden were selected from a group of Finnish twins of the Finnish Twin Cohort Study. Food consumption data were collected by means of diet history interviews. The estimated intakes of 30 nutrients calculated on the basis of PC-kost were compared to the corresponding estimates calculated on the basis of Nutrica. The calculation procedures were standardised.

No statistically significant differences were observed in the mean intakes of energy, total fat, saturated fat, carbohydrates, dietary fibre, alcohol, cholesterol, vitamin A, retinol, beta-carotene, vitamin D, alpha-tocopherol, riboflavin, niacin, vitamin B(12), vitamin C or phosphorus. PC-kost yielded a 20% higher intake (NS) for vitamin D and 23% higher intake (P<0.001) for thiamine than Nutrica, which is mainly attributed to the differences in the enrichment of foodstuffs between Sweden and Finland. Conversely, PC-kost yielded 53% lower values (P<0.001) for selenium than Nutrica, owing to the increased selenium content in many Finnish foodstuffs as a result of the addition of selenium to fertilisers. Statistically significant differences were found for protein, monounsaturated fatty acids, vitamin B(6), iron and sodium (5-9% higher values from PC-kost) and for polyunsaturated fatty acids, folate, zinc, calcium, magnesium and potassium (4-10% lower values from PC-kost). The variation in the intake of these nutrients between the two methods may be explained by the differences in foodstuff-specific nutrient values (eg product formulations), or differences in the sources of data, recipes or calculation procedures. The correlation coefficient was > or =0.81 for most nutrients. At least 85% of the subjects in each PC-kost quintile were classified into the same or adjacent Nutrica quintile for all nutrients.

Our results indicate that, for a dominant part of the examined nutrients, the estimated intakes calculated by means of standardised procedures using the PC-kost and Nutrica databases are comparable between Sweden and Finland. Differences observed for some nutrients reflect either actual differences in foods between the two countries or methodological differences in the assessment of nutrient intakes.