Arsenic in seafoods: Human health aspects and regulations

Abstract

Some countries, for reasons of human health, set standards to control the amount of metals and metalloids in seafoods. Such standards usually take the form of maximum permissible concentrations (MPCs). Many metallic elements are naturally present in marine organisms, and to accommodate traditional foodstuffs, standards must take account of what levels should be considered normal, and this can differ from one type of animal to another. For example the Australian MPC for cadmium in fish is 0.2 mg kg -j but for molluscs is 2.0 mg kg -~ (Anon., 1991). This reflects what are considered normal concentrations for fish and molluscs, and also takes account of the average contributions of both to the diet of Australians. Presumably the regulations of other countries were formulated in a similar manner, and must chiefly be designed to prevent contaminated food from entering the diet. For most metallic elements that do not form stable chemical bonds with carbon under physiological conditions, it is enough to set a standard for the metal itself. This is true, for example, for cadmium. Although organic derivatives of cadmium are known, they are not stable under normal conditions and cadmium in the human gut tract is likely to be present essentially as the Cd ++ ion. Although the absorption and toxicity of this cadmium may be affected by other substances present, it is nevertheless realistic to set standards based upon the expectation that Cd ++ is the sole chemical species being covered. This is not true for mercury; a sample may contain inorganic mercury, or methylmercury, or other organic forms of mercury, or, of course, combinations of these. A wealth of analytical data collected widely over many years has shown that the vast majority of fish contain most of this mercury in the toxic methylmercury t'orm. Presumably the health regulations of the various countries take note of this as one of a number of factors, although few make specific mention of it. The situation with arsenic is more complicated. Chemists and toxicologists have long known that the toxicity of arsenicals is highly dependent upon the nature of the compounds, particularly upon the valency state of the arsenic atom (Vallee et al., 1960; Frost, 1967; Penrose 1974). In general, compounds containing trivalent arsenic are much more toxic than those with a pentavalent arsenic atom. There is also a commonly-held notion that inorganic arsenic is more toxic than organic arsenic. Quite possibly such an idea has grown from a knowledge that organic arsenicals were used as medicines, and from early observations, initially perhaps those of Chapman (1926) and Coulson et al. (1935), that seafoods contained organic arsenic compounds and did not appear to be in any way harmful, whereas it was well known that inorganic compounds of arsenic, such as arsenous oxide, were very poisonous. The valency of the arsenic atom is, however, a more important factor in determining toxicity than the organic or inorganic nature of the compounds; trivalent organic arsenic compounds tend to be highly toxic. For inorganic arsenic, salts of arsenic acid (arsenates) with arsenic in the pentavalent state are less toxic than salts with arsenic in the trivalent state (arsenites) (Penrose, 1974). However, some reduction of arsenate to arsenite might occur within the mammalian body (Vahter & Envall, 1983) and it would be unwise to disregard the possible eventual toxicity of inorganic arsenic ingested in either valency state. Pentavalent organic arsenic compounds appear to be of low toxicity or to be non-toxic, at least as far as toxicity resulting from the presence of the arsenic atom is concerned. This is particularly so if the arsenic atom cannot be reduced to the trivalent state without the severing of an arsenic-carbon bound. These bonds are stable and there have been no reports to our knowledge of breakage, resulting, for example, in demethylation, occurring in an animal body.