 |
INTRODUCTION |
Similarities in ethnic background, prevalence of noncommunicable diseases, food habits and other lifestyle-related factors in the Nordic countries have generated a number of initiatives for collaboration and harmonization in food and health issues. Since 1980, the Nordic countries have established common official nutrition recommendations and have harmonized regulations and monitoring on several questions regarding food toxicology and risk assessment. However, only occasionally have there been formal contacts between nutritionists and toxicologists in the evaluation of safe levels of intake of essential nutrients. In connection with the 3rd revision of the Nordic Nutrition Recommendations (NNR)2 (Standing Nordic Committee on Foods 1996), a working group was established for the risk assessment of selected essential trace elements. The results of this evaluation (Nordic Working Group 1995) were used to establish values for upper limits of intake for iron, zinc, iodine and selenium in the NNR (Table 1). During this work, it become apparent that neither the criteria used for toxicological evaluation nor the safety factors commonly used in risk assessment of food additives and contaminants were applicable to essential trace elements.
View this table:
[in this window]
[in a new window]
|
Table 1.
Recommended and upper limit for daily intake of selected trace elements in the Nordic Nutrition Recommendations 1996 (Standing Nordic Committee on Food 1996) together with estimates of average intake per 10 MJ in the
Nordic countries
|
|
| |
IODINE |
Iodination of table salt is allowed in Sweden, Finland and Norway and, together with enrichment of cattle feed, contributes to the existence of average iodine intakes above recommended intakes. A recent dietary survey in Denmark revealed a median iodine intake of 100 µg/d in women and 130 µg/d in men (Andersen et al. 1996
). In Iceland, the intake of iodine is high (~300 µg/d) due to high fish consumption (Steingrímsdóttir et al. 1991). Occasionally, dietary iodine intakes on the order of 1-10 mg/d have been reported as a result of consumption of seaweed products (Skare and Frey 1980).
The upper limit of iodine intake is based on observations of thyroid hypo- and hyperfunction at high intakes. In most healthy subjects, intakes of ~1 mg/d are well tolerated. However, there seem to be subgroups in the population who, through unrevealed biochemical mechanisms, develop goiter and/or hypothyroidism or excessive activity of the thyroid gland at intakes on the order of 300-1000 µg/d (Bürgi et al. 1982). Despite indications of negative effects in individual subjects at lower intakes, the upper limit was set at 1000 µg/d. Even at that level, the safety factor in relation to typical dietary intakes is low especially among residents of Iceland and individuals consuming seaweed products.
| |
SELENIUM |
Geochemically, the Nordic countries represent an area low in selenium. In Denmark and Sweden, residents subsisting on nationally produced grains consume 40-50 µg/d. Residents of Finland consume 100-200 µg/d as a result of the use of selenium-enriched fertilizers. Similar intakes are reported from Norway and Iceland because of the importation of high selenium grains.
Toxic effects of high doses of selenium are well known in both domestic animals and humans. Clinical signs of selenosis (hair or nail losses, nail abnormalities, skin lesions or changes in peripheral nerves) have been observed at intakes of 1200-5000 µg/d (Yang et al. 1993). Biochemical changes (prolonged plasma prothrombin time and decreased concentrations of blood glutathione) have been reported at dietary intakes exceeding 750-850 µg/d (Yang et al. 1989a and 1989b). Marginal hematological changes and a borderline increase in serum alanine aminotransferase (ALAT) have been reported in two subjects as a consequence of intake of selenium-containing yeast at doses of 200 and 400 µg/d (total intake ~350-600 µg/d) for 18 mo (Schrauzer and White 1978). A long-term study of subjects in a high selenium area with an average intake of 239 µg/d showed no association between selenium intake and prothrombin time but a positive association with ALAT activity (Longnecker et al. 1991).
The working group recognized that the clinical relevance of these biochemical changes is uncertain but suggested that intakes causing these effects could be taken as lowest-observable-adverse-effect levels (LOAEL) for selenium, with 300 µg/d set as the upper intake limit. The safety margin in relation to current dietary selenium intakes in Finland, Norway and Iceland is low, and the upper limit could be exceeded easily by intake of available selenium supplements.
| |
IRON |
Current iron intakes in the Nordic countries are 10-12 mg/10 MJ and the prevalence of iron deficiency (serum ferritin <12-15 µg/L) is high. Recent investigations reported a 10-22% prevalence of iron deficiency among women of child-bearing age and 6-40% among teenage girls (Standing Nordic Committee on Foods 1996). Thus, there is a case for iron enrichment, supplementation or other actions to improve the iron status in relatively large segments of the population.
Chronic iron overload, seen in patients with hemochromatosis, leads to liver fibrosis, cirrhosis and hepatocellular carcinoma (Bonkovski 1991). In addition, diabetes, hypogonadism and cardiac disease are observed. Results from the Nordic countries suggest that the prevalence of homozygotes for hemochromatosis is 0.05-0.5% and of heterozygotes 8-14% (Bacon 1992, Nordic Working Group 1995). There are no reports of increased incidence of disease among heterozygotes. However, significantly higher hepatic iron concentrations have been reported (Bassett et al. 1981, Olsson et al. 1987), suggesting impaired iron regulation also in heterozygotes.
Iron overload due to excessive dietary iron intake derived from the use of iron utensils for cooking and preparation of beer has been reported among African Bantu (FAO/WHO 1983). Iron exposure was estimated to be 50-100 mg/d. Autopsies revealed high hepatic iron stores in a substantial part of the population and a high prevalence of liver cirrhosis.
Few reports are available regarding long-term iron therapy and iron overload. High ferritin concentrations after prolonged iron supplementation have been reported in a few cases (Nordic Working Group 1995).
The potential risk of iron overload is related to ferrous ion catalyzing the formation of hydroxyl radicals, resulting in oxidative damage to membranes, cell organelles and the genome. In addition, a direct effect of iron on collagen biosynthesis has been proposed (Bacon and Britton 1989). Iron bound to transferrin or ferritin may be released at low pH or when exposed to organic free radicals or superoxide (Aust et al. 1993). It has been suggested that there is a small body pool of low molecular chelated iron with a potential to promote oxidative damage.
Because oxidative damage is involved in the pathogenesis of coronary heart disease (CHD) and cancer, it has been suggested that a high iron intake and high iron stores could increase the risk for these diseases. Subchronic and chronic studies in laboratory animals have shown that iron doses resulting in iron overload potentiate the effect of known carcinogens. Results from studies using isolated tissues and organs from experimental animals suggest that iron overload increases the susceptibility to oxygen reperfusion damage (Nordic Working Group 1995). Results from epidemiologic studies in humans are inconsistent and inconclusive. In a Finnish study of 1931 middle-aged men, a 2.2-fold higher risk (95% confidence interval 1.2-4.0, P < 0.01) of acute myocardial infarction was observed in men with serum ferritin concentrations > 200 µg/L (Salonen et al. 1992). Other studies have not been able to find an association between iron status and risk of CHD (Sempos et al. 1996). With regard to cancer risk, some epidemiologic studies suggest a positive association between indices of iron status and risk of development and mortality of cancer (Nordic Working Group 1995). The interpretation of these epidemiologic observations is confounded by the use of unreliable indices of iron status and lack of identification of subjects with hemochromatosis; given current knowledge, it is not possible to identify a daily iron intake that would represent a health risk.
Excessive iron intake could result in negative interactions with other trace elements. Inorganic iron added to aqueous solutions of zinc in molar ratios of Fe/Zn of 2:1-25:1 reduced zinc absorption, whereas the same amounts of iron added to a test meal had no negative effect (Sandström et al. 1985, Valberg et al. 1984). Decreased serum zinc concentrations have been reported after iron supplementation with doses > 60 mg/d (Breskin et al. 1983, Hambidge et al. 1983 and 1987), but also at a level of iron supplementation of only 18 mg/d in pregnant teenagers (Dawson et al. 1989). In contrast, Sheldon et al. (1985) reported no effect of supplementary iron on serum zinc concentration during pregnancy.
Interactions between iron and manganese have been demonstrated in experimental animals as well as in humans. An increased dietary manganese concentration impairs iron absorption in a manner suggesting that the body cannot distinguish between iron and manganese ions in the absorptive process (Rossander-Hultén et al. 1991). Iron fortification of food seems to have no effect or only moderately negative effects on manganese absorption (Davidsson et al. 1989 and 1991), whereas it is not known whether higher iron intakes in the form of supplements have more pronounced effects.
The upper limit of iron intake in the NNR was based on observations from subjects with primary hemochromatosis and to some extent on reports of hemosiderosis among the Bantu in South Africa, observations that demonstrated an increased risk of disease or other adverse effects in states of iron overload. However, none of these studies could be used to quantify the intake leading to an increased risk. Therefore, a calculation made by Fairbanks and Beutler (1988) was used to identify the hypothetical iron intake not leading to iron overload. These calculations take into account the effect of dose and iron stores on iron absorption as well as the effect of iron stores on excretion. The cut-off values indicative of iron overload suggested by an American expert group (Expert Scientific Working Group 1985) were used in this evaluation. The algorithms for a woman of fertile age were used as a basis for the calculation. Under the assumptions given, a daily intake of 60 mg iron can be estimated to produce a ferritin concentration of 140 µg/L after 5 y, a value just below the suggested cut-off value of 150 µg/L. At daily doses of 120 mg, this concentration is theoretically reached after 2 y. The results from studies on the interaction with other trace elements suggest that the observed "no-adverse-effect level" is lower than the proposed 60 mg. However, systematic dose-response studies that would allow the application a more restrictive value for the upper limit of intake of iron are lacking.
| |
COPPER |
As with iron, copper acts as an efficient in vitro prooxidant; it has been suggested that increased plasma copper concentrations are associated with an increased risk of coronary heart disease (Salonen et al. 1991). It is unclear, however, to what extent increased copper concentrations reflected a high dietary intake of copper or indicated inflammatory processes. Gastrointestinal disturbances such as nausea, vomiting and abdominal cramps have been reported at daily copper intakes ranging from 2 to 32 mg from contaminated water. Infants and children seem to be especially sensitive to high copper intakes. Swedish studies suggested that diarrhea in young children could be due to high copper contents in formula prepared with copper-contaminated water from copper-containing domestic piping (Nordic Working Group 1995). Cases of liver disease in children exposed to high copper contents in drinking water (Müller-Höcker et al. 1987 and 1988) support the concern regarding children as a risk group for toxic copper intakes. The working group could not find sufficient documentation for an upper limit of intake of copper but recommended that the copper concentration in drinking water should not exceed 2 mg/L.
| |
CONCLUDING COMMENTS |
Different criteria were used for the individual trace elements to set an upper limit for intake in the NNR. Because of a lack of sufficient data, pragmatic decisions had to be made for several of the elements. A conventional toxicological approach with a reasonable safety factor applied to assumed "no-observed-adverse-effect levels" was not possible for any of the elements evaluated because the derived values would have overlapped the values for recommended intakes for essentiality. There is an urgent need for systematic dose-response studies of potential interactions between trace elements as well as for studies on the effect of chronic exposure to elevated intakes of trace elements.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Abstract
Introduction
Abstract
Essential versus Toxic Levels of Intake (Oskarsson, A., ed.), Nord 1995:18. Nordic Council of Ministers, Copenhagen, Denmark.
