of common methods of collecting data and accounting for use over time. The disadvantage of using disappearance data to estimate nutrient intake is that it overestimates intake because it fails to capture food losses and wastage after the nutrient enters the food system (e.g., cooking and processing losses).

Salt disappearance data can be used to estimate time trend patterns in the availability of sodium for human consumption. The Salt Institute posts information on its website about food-grade salt sales in the United States.¹⁵ These data are most useful if the tonnage of salt is converted to milligrams of sodium. With changing population numbers over time, it is also useful to convert annual results to per capita values. The annual per capita sodium disappearance numbers from 1978 through 2008 derived from data on salt disappearance are illustrated in Figure 2-12.

The salt disappearance data show a steady increase in per capita availability between 1983 and 1998. More recently, values appear to be leveling off or decreasing slightly. The peak levels in 1998 indicate that approximately 5,700 mg of sodium were available per person per day. The extent to which the disappearance values are an overestimation of actual intake is unknown but the fact that they are more than double the Dietary Guidelines for Americans level of < 2,300 mg/d sodium suggests that salt availability is in excess of public health goals for sodium. Moreover, given that the major advantage of disappearance data is the trend pattern that they reveal, the disappearance data in Figure 2-12 do not show a sustained reduction in response to the sodium-related public health initiatives identified in Tables 2-1 to 2-3. Although the pattern of use over time suggests that early educational and program initiatives carried out in the 1980s were associated with a reduction in salt use, subsequent programs—including the implementation in 1993 of mandatory declaration of sodium content on all food labels and multiple calls since 1969 for food processors to reduce the sodium content of foods—appear to have had little or no impact on salt availability for human use.

Since 1971, NHANES has provided estimates of individuals’ nutrient intakes from a nationally representative sample of the U.S. population. These estimates are based on 24-hour recalls. As shown in Figure 2-13, the trends in sodium intake between the 1971–1974 and 2005–2006 surveys are shown for three life stage groups. Similar patterns were seen across other life stage groups (Briefel and Johnson, 2004; see Chapter 5).