Comparability of household and individual food consumption data--evidence from Sweden

Comparison of household and individual food consumption.

Combined household and individual food consumption survey carried out in Sweden in 1989. A random sample of 3000 subjects aged 0-74 years, the household to which the subject belonged constituted the household unit. Each household recorded all the foods it purchased over a 4-week period, except food eaten outside the home. For the selected subject, excluding children <1 year old, food intake was obtained with a simplified 7-day record. Food consumption from the two data sets was compared for the whole sample and for one-person households, respectively.

Reasonable agreement (+/-20%) was seen for many major foods including cereal products, milk, cheese, meat and meat products, and fish, and after correction for inedible part, fruit and vegetables. Purchases were lower for sweet bakery products, alcoholic beverages and potatoes, while the opposite was seen for oils and fats, cream and sugar. For one-person households the two data sets showed similar differences in consumption patterns between men and women. The gross differences observed can be explained by factors like home baking, eating out and recorded level of processing, e.g. as raw food including inedible part or as a prepared dish.

The results show reasonable agreement between purchased and eaten amounts for aggregated food groups but marked differences were seen for some important individual items. Household-based consumption data are useful for many purposes provided the limitations of the data are accounted for.