Food supply versus household survey data: nutrient consumption trends for Spain, 1958-1988

Various methods of estimating food consumption, such as food balance sheets (FBS) and household surveys (HS), have been developed over the years and have been used to inform, monitor and evaluate nutrition policies. Because these methods vary in their objectives and data collection procedures, the objective of this study has been to elaborate FBS data for Spain and to study the consistency of fat, carbohydrate and protein intake trends, as measured by FBS and HS, for the period 1958 to 1988. Food balance sheets were elaborated by the authors according to the methodology of FAO using all available data sources for the 1958-1988 period. This data considered every major food item contributing to the total energy intake of the spanish population. Household survey data were taken from three similar national household budget surveys carried out on a representative sample of the Spanish population in 1958, 1964-1965, and 1980-1981. Estimates of food consumption were transformed into macronutrient intake by applying standard food tables. When macronutrient intake were expressed in absolute amounts, an unexpected finding was the tendency of the household surveys to overestimate food balance sheet data for fat, and to a lesser extent protein and carbohydrate, intake during the first years in the series. Also, the slopes of the trends of macronutrient intake were significantly (p < 0.05) greater for food balance sheets than for household survey data, specially for fat. When macronutrient intake were expressed as percent of total energy, differences between the two types of data tended to diminish and heterogeneity of slopes disappeared. We conclude that household survey and food supply data provide partially different information on macronutrient intake trends in the Spanish population for the period 1958-1988. The discrepancy is particularly noticeable for fat intake and when data are expressed in absolute amounts.