Interpretation of dietary change in the United States: fat as an indicator

Measurement of dietary trends poses methodologic issues of great complexity and intellectural challenge. Confidence in the ability to draw inferences about trends increases when the results of various measurement methods point in the same direction, are of comparable magnitude and are consistent with one another as well as with external sources of related data. This paper illustrates these issues by reviewing inconsistencies in available data sources since the early 1970s on intake of fat in the United States.