Identification of a No-Observed-Adverse-Effect Level (NOAEL) and a Lowest-Observed-Adverse-Effects Level (LOAEL). The lowest blood level of selenium measured in the five subjects at initial examination was 13.3 µmol/L (105 µg/dL), corresponding to a selenium intake of 913 µg (12 µmol)/day (range: 913 to 1,907 µg [12 to 24 µmol]/day). The average blood selenium level was 16.9 µmol/L (135 µg/dL). At the time of reexamination in 1992, all five patients were described as recovered from selenium poisoning, although their fingernails reportedly appeared brittle. The mean blood selenium level had decreased to 12.3 µmol/L (97 µg/dL), corresponding to a selenium intake of about 800 µg (10 µmol)/day (range 654 to 952 µg [8.3 to 12 µmol]/day). The lower limit of the 95 percent confidence interval was 600 µg (7.6 µmol)/day.

Yang and Zhou (1994) therefore suggested that 913 µg (12 µmol)/day of selenium intake represents an individual marginal toxic daily selenium intake or LOAEL. They further suggested that the mean selenium intake upon reexamination (800 µg [10 µmol]/day), represented a NOAEL, while 600 µg (7.6 µmol)/day of selenium intake was the lower 95 percent confidence limit for the NOAEL. These values appear reasonable, although the number of subjects was small. Nevertheless, the LOAEL for selenosis in this small data set appears to be representative of the larger data set, and the reexamination of the subjects provides valuable dose-response data. Uncertainty occurs because of the smallness of the data set and because the Chinese subjects may not be typical (e.g., they may be more or less sensitive to selenium than other populations).

Longnecker et al. (1991) studied 142 ranchers, both men and women, from eastern Wyoming and western South Dakota who were recruited to participate and were suspected of having high selenium intakes based on the occurrence of selenosis in livestock raised in that region. Average selenium intake was 239 µg (3 µmol)/day. Dietary intake and selenium in body tissues (whole blood, serum, urine, toenails) were highly correlated. Blood selenium concentrations in this western U.S. population were related to selenium intake in a similar manner to that found in the Chinese studies, presumably because the form of selenium ingested was selenomethionine. No evidence of selenosis was reported, nor were there any alterations in enzyme activities, prothrombin times, or hematology that could be attributed to selenium intake. The highest selenium intake in the study was 724 µg (9 µmol)/day.

It thus appears that a UL based on the Chinese studies is protective for the population in the United States and Canada. Therefore a NOAEL of 800 µg (10 µmol)/day is selected.