Heavy metal exposure from foods.

The Food and Drug Administration has a continuing program of monitoring foods for their content of lead, cadmium, mercury, zinc, arsenic, and selenium to determine trends of increasing or decreasing levels. The monitoring protocol is that of the Total Diet Study, in which "market baskets" of typical foods and beverages consumed by 15- to 20-year-old American males are collected in various geographical locations at regular intervals during the year, divided into food classes, composited, and analyzed. Cadmium has the most widespread distribution of the six heavy metals and mercury the most limited. The analytical values for lead may be underestimated because of limitations of the methodology; these do not apply to the other five elements. A tabulation by year shows that the levels of these elements in foods do not vary significantly from one year to the next. Average intakes of lead, cadmium, and mercury are below the WHO/FAO tolerable intakes for adults; such tolerable intakes have not been established for arsenic and selenium. Increases in concentrations of these elements in foods would be considered undesirable, however.

Determination of the heavy metal content of foods has been carried out by various laboratories of the Food and Drug Administration for a number of years. The data discussed were obtained from the Total Diet Survey program, or as it is better known, the Market Basket Survey. The program was originally designed to monitor the levels of various pesticides in foods and has been ongoing since 1965. In the last few years it has been expanded to include analyses for six heavy metals: lead, cadmium, mercury, zinc, arsenic, and selenium.t Lead, zinc, and selenium have been determined only since 1972, whereas data from earlier years are also available for the other three metals. A major purpose of this program has been to determine the levels of heavy metals in foods, or more specifically, to determine trends of increasing or decreasing levels of metals in various foods or food classes.
The foods included in the Total Diet Study were determined by using 1965 diet survey data of the United States Department of Agriculture which indicate the food consumed by 15-to 20-yr-old *Food and Drug Administration, 200 C St. S.W., Washington, D.C. 20204. tIn this context, the term "heavy metals" is used in a colloquial sense and includes the toxic nonmetals arsenic and selenium.
American males (1). Food consumption of this group was selected for study because the members of the group consume the largest number of calories of any age-sex group in the population, and the largest consumers of food generally have the highest exposure to food contaminants. However, it is recognized that there are possible exceptions to this generality if a food contaminant were present in only a few types of foods.
By using this data base it has been found that over a 14-day period the average diet of young The foods are not analyzed individually but as part of a composite food class, as shown in Table 1. The actual composition of the garden fruits class is shown in Table 2. In all cases portions of the individual samples are retained. In actual practice, more than enough of each food item is collected for the requirements of the 14-day composite. If the level of a contaminant in a composite is unusually high the individual components of the composite will be analyzed to determine the source. Although it may be desirable to analyze all the foods individually, the costs would be very greatly increased. Qualitative evaluation of the data reveals that the pattern of distribution in foods differs substantially among the various metals. Cadmium has the most widespread distribution (Table 3). Specific sources of high concentrations of cadmium are shellfish and some vegetables such as spinach; however, when both food consumption and concentration of the metal are considered, a number of composites provide substantial portions of the cadmium intake. Among these food classes, cereals and grains provided the greatest percentage of total intake. Lead is also found in a large number of foods. However, fruits and vegetables are the most important sources. While they constitute only 23% of the total diet by weight, they account for 78% of the total lead intake (Table 3). Zinc was present in all the food classes examined. However, dairy products, meat-fish-poultry, and grains and cereals provide 77% of the zinc (Table 3), even though they account for only about half the diet by weight.
Arsenic, which is reported as As20s, has a narrower distribution. In recent years, arsenic occurred most frequently in the meat-fish-poultry composites. Shellfish have been reported to contain generally higher levels, but they are not significant components of the average diet. Patterns of food consumption are such that meat-fish-poultry, dairy products, and grains and cereals are important dietary sources. Of the total arsenic in the diets studied in this program, 92% came from these three food classes (Table 4). With regard to selenium, essentially all selenium in the diet is attributable to the meat-fish-poultry group and the grain-cereal group. These two food classes provide over 99% of the selenium intake (Table 4).
Mercury has the most fimited distribution of the metals studied. Virtually all dietary intake of mercury was found in the meat-fish-poultry class and in fish within that class where followup studies were performed (Table 4). This observation is further supported by Simpson et al. (2) in a report -4 Environmental. Health-Perspeeives. of a study on fish, other staple foods, and total diet composites. The report indicated that, except for fish, mercury levels in foods are generally so low as to be undetected by the usual flameless atomic absorption techniques; neutron activation analysis was necessary to detect any mercury present. The Market Basket surveys as currently performed give a good indication of general trends of distribution of the metals in foods. However, in some cases there may be greater intakes from some food classes than currently apparent because of the detection limits of the chemical methods employed. This problem becomes particularly apparent in the food classes that constitute a large percent of the total diet but may contain low concentrations of some of the metals: dairy products, meat-fish-poultry and beverages.
Detailed data on the findings of heavy metals in various food classes are presented in Tables 5 and 6. In evaluating these data it has been recognized that the use of food composites creates difficulties in determining actual total quantities of metals present in foods. For example, the concentration of lead in many-ofthe composites is very-close to the detection limit of the method. Thus, analyses of many of the composites will yield the analytical report of trace or zero. How these traces or zeros are handled in calculations will substantially affect the magnitude of the value reported for total intake. This problem has been discussed in depth in a previous publication (3). A wide range of total lead intakes can be calculated based on various analytical assumptions (Table 7). Among the metals included in the Total Diet Study, this problem is most severe for lead. Cadmium, arsenic, and selenium present far fewer difficulties in this regard because of lower levels of detectability in relation to the levels that occur. The chemical methodology for zinc and mercury is sufficiently reliable at the levels determined that the estimated intakes for these metals are considered to be very accurate approximations of the actual values. Table 8 presents the actual values found in the 1973 Total Diet Survey and the effects of various assumptions on them. Although total exposure to lead may be underestimated, this is not as significant a problem for most of the metals determined in this study. Table 9 shows yearly values in the metal content of foods, as obtained from the Total Diet Surveys. Although the total dietary lead content is undoubtedly underestimated, the chemical methodology and handling of traces and zeros in calculations has remained constant, and the same is true for other metals. Thus comparisons of yearly changes are meaningful. These data, shown in Ta-   In evaluating the toxicological significance of the various metals, the estimated intake from the Total Diet Study has been compared with the WHO/FAO provisional recommendations of tolerable intakes (4) for adults (Table 10). Concerning lead, two points are apparent. Because of underestimation of actual lead intake the per cent of the recommended level is far higher than the 14% which is calculated. However, on the basis of data presented in an earlier paper from our group it is known that the average lead intake is well below For cadmium and mercury the percent of tolerable intake is considered a good estimate. Cadmium is close enough to the tolerable intake so that further increases in the cadmium content of foods should be avoided. Thus far WHO/FAO tolerable intakes for arsenic and selenium have not been established. Selenium and zinc are both essential nutrients. The National Research Council of the USA recommends an intake of 15 mg zinc/day for adults (5). The teenage boy who consumes a large amount of food can easily meet this requirement. However, the recommended dietary allowances for pregnant and lactating women are 20 and 25 mg zinc per day, respectively. With lower calorie intakes and a higher requirement, foods must be carefully selected in order to avoid the problem of an undesirably low zinc intake. National Research Council recommendations on selenium intake have not been set (5).
In summary, for the metals included in the Total Diet Study it is generally considered that while the food supply contains less than tolerable intakes of those toxic heavy metals for which recommendations exist, any increases in these trace metal concentrations of food are undesirable. ;68 --