点击显示 收起
1 From the Institute of Nutrition and Food Technology, University of Chile, Santiago.
2 Supported by Fondecyt Chile grant 1000852 and by an unrestricted research grant from the Copper Risk Assessment Research Program in Chile, managed by the Chilean Center for Mining & Metallurgy Research and the International Copper Association. 3 Address reprint requests to M Araya, Institute of Nutrition and Food Technology, Macul 5540, Santiago 11, Chile. E-mail: maraya{at}uec.inta.uchile.cl.
ABSTRACT
Background: Mild and moderate effects of marginally low and marginally high copper exposure are poorly understood in humans.
Objective: The objective was to assess acute gastrointestinal effects and blood markers of copper status in apparently healthy adults who underwent controlled copper exposure for 2 mo.
Design: This was a 2-mo, randomized, controlled, double-blind study of 1365 apparently healthy adults in whom acute gastrointestinal symptoms (nausea, vomiting, diarrhea, and abdominal pain) were assessed as responses to copper exposure (<0.01, 2, 4, or 6 mg/L water). Blood markers were measured in 240 participants at the end of the survey. Subjects with anemia, inflammation, or infection were excluded. Serum and erythrocyte copper, peripheral mononuclear cell copper, serum ceruloplasmin, the nonceruloplasmin bound copper fraction, superoxide dismutase activity, hemoglobin, mean corpuscular volume, serum ferritin, and liver enzyme activities were measured.
Results: The percentage of subjects reporting gastrointestinal symptoms was higher in the 6-mg Cu group than in the <0.01-mg Cu group (P < 0.02). One hundred ninety-five subjects fulfilled the inclusion criteria for the blood studies. Although a significant relation between copper intake and total gastrointestinal symptoms was observed, no relation was found between copper intake or reported symptoms and copper-load variables.
Conclusions: Gastrointestinal symptoms increased significantly in response to 6 mg Cu/L water. No detectable changes were observed in indicators of copper status, which suggests competent homeostatic regulation. The results of liver function tests remained normal in all subjects. The lack of change in superoxide dismutase activity supports the Food and Nutrition Board’s latest recommendation of 0.9 mg Cu/d for adults.
Key Words: Serum copper • ceruloplasmin • superoxide dismutase • SOD • nonceruloplasmin copper fraction • homeostasis • gastrointestinal symptoms
INTRODUCTION
The persistence of health problems related to micronutrient deficiencies in large groups of the world’s population brought attention to the relevance of microminerals in human health. Of these microminerals, copper is an interesting element because it is both essential and toxic to humans, depending on the dose ingested. Adverse effects associated with a severe lack or excess of copper are well described in genetic conditions (Menkes and Wilson diseases, respectively). Acquired copper deficiency is described in children recovering from malnutrition, in low-birth-weight neonates, and in patients receiving copper-free total parenteral nutrition. Effects of high copper exposure in persons that do not carry specific mutations that render them more susceptible are less clear. One of the difficulties in defining the milder effects of copper deficiency or excess is that limits of homeostatic regulation are not known. Identification of the early biologically relevant effects of copper excess and copper deficiency and the markers of changes in copper status are crucial to defining successful and safe nutritional interventions with micronutrients in vulnerable and nonvulnerable groups.
The extent to which classic markers of copper status may change within the limits of homeostatic regulation has not been systematically assessed. Serum copper and ceruloplasmin concentrations and superoxide dismutase (SOD) activity were shown to decrease in some persons who received 12.4–16.2 µmol Cu/d (0.79–1.03 mg Cu/d) (1–3). No potential early markers of copper excess have been identified. Studies by Olivares et al (4) showed that infants who received 2.8 µmol Cu/d (180 µg·kg-1·d-1) during the first year of life remained asymptomatic, grew normally, and had liver function within the normal range. Although the nonceruloplasmin copper fraction correlated positively with serum copper values, no significant changes in serum copper, ceruloplasmin, or SOD or metallothionein in erythrocytes were observed in the infants throughout the study. Whether these results relate to age is not clear.
This study was planned to explore the early effects of copper in an apparently healthy adult population that underwent a period of controlled exposure to copper at concentrations that differed 10-fold. The objectives of the study were to assess potential changes in blood indicators of copper status at these levels of exposure and to establish whether reports of acute gastrointestinal symptoms were related to changes detected in the blood indicators measured.
SUBJECTS AND METHODS
Design
The study consisted of a 2-mo, randomized, controlled, double-blind study of 1365 apparently healthy adult participants in whom acute gastrointestinal symptoms (nausea, vomiting, diarrhea, and abdominal pain) to copper exposure in water were surveyed. The protocol was carried out in suburban Santiago, Chile, where previous studies (5) showed that the average daily copper intake is 14.2 ± 12.6 µmol/d (0.9 ± 0.8 mg/d). Tap water in Santiago provides a mean of 0.2 µmol Cu/L (0.01 mg Cu/L) (Empresa Metropolitana de Obras Sanitarias, Santiago, Chile). The basal dietary copper intake of the subjects was estimated to be 14.3 µmol/d (0.91 mg/d); thus, it was estimated that the long-term copper status of the population would be borderline low. Families participating in this survey prepared the water to be consumed at home on a daily basis, using tap water and a stock solution provided by the researchers. Thus, 10 L water was prepared daily in 20-L containers and was used as needed during the day by all adult participants in the household. The subjects drank the water or used it to prepare fluids to be consumed at home (eg, tea, coffee, homemade infusions, and soups). Depending on the stock solution used for water preparation, families were exposed to <0.2, 31.5, 63.0, or 94.4 µmol Cu/L (<0.01, 2, 4, or 6 mg Cu/L, respectively) as copper sulfate; final concentrations were verified by atomic absorption spectrometry in a weekly sample from each household. Any gastrointestinal symptoms experienced daily by the subjects were recorded in diaries, which were reviewed at the subjects’ homes every other day by trained fieldworkers. The subjects chose from a list of symptoms defined as outcomes of the study (nausea, vomiting, diarrhea, and abdominal pain) and other symptoms that were used to distract the participant from the variables of interest. The diary was based on results obtained from previous studies of the effects of copper on health and was validated before the beginning of this study.
During the survey, individual mean daily fluid consumption was 1.5 L. Thus, because the basal daily dietary copper intake from foods was estimated to be 14.2 µmol/d (0.9 mg Cu/d) in families who consumed <0.2 µmol/L (<0.01 mg Cu/L) in water, the mean daily copper intake was 14.3 µmol/d (0.91 mg/d); individual subjects who consumed 94.4 µmol Cu/L (6 mg Cu/L) in water had a mean daily copper intake of 155.8 µmol Cu/d (9.9 mg Cu/d). These values equate to mean individual daily intakes of 0.2 and 2.4 µmol·kg-1·d-1 (15 and 150 µg·kg-1·d-1), respectively. The copper concentrations to be tested were as follows: <0.01 mg/L, which served as the control; 2 mg/L, which represented the current provisional guideline for copper in drinking water established by the World Health Organization (6); 4 mg/L, the concentration at which reports of gastrointestinal symptoms became significantly higher in previous clinical trials (7, 8); and 6 mg/L, the concentration at which vomiting was first reported in the clinical trials. This basal survey was intended to assess gastrointestinal symptoms and copper consumption (in fluids) under conditions as close to real life as possible. The acute gastrointestinal symptoms reported by the 1365 participants during the 2-mo health survey will be reported in detail elsewhere.
Subjects
Of the 1365 participants, a total of 240 (60 from each group that consumed <0.01, 2, 4, or 6 mg Cu/L during the 2-mo survey) provided a blood sample. The subjects were aged 20–55 y, and the sample was balanced by sex. All participants signed an appropriate informed consent form before participating in the survey and before providing blood samples. The protocol was approved by the Institutional Review Board, Institute of Nutrition and Food Technology, University of Chile.
Variables and procedures
The assessment of copper status included the measurement of copper concentrations in serum and in erythrocytes by atomic absorption spectrometry (model 2280; Perkin-Elmer, Norwalk, CT). In addition, the intracellular copper content was determined in mononuclear cells (1 x 106) by total reflection X-ray fluorescence spectroscopy, as previously described (9). Ceruloplasmin concentrations were measured by nephelometry (Array Protein System; Beckman Instruments Inc, Brea, CA). The pool of nonceruloplasmin-bound copper was calculated on the basis of the assumption that the molecular weight of ceruloplasmin is 132000 and that each molecule binds 6 copper atoms (10).
SOD activity was determined in erythrocytes by using a commercial kit (Bioxytech SOD-525 Assay; OXIS International Inc, Portland, OR). The calibration curve was produced with purified SOD (EC 1.15.1.1) from bovine liver (Sigma, St Louis), and the results are expressed as units of SOD/mg hemoglobin.
Before the blood samples were collected, a questionnaire was used to exclude from sampling the subjects who had any clinical manifestations suggestive of infection or inflammatory processes. The C-reactive protein concentration was used as an additional criterion to exclude from sampling the subjects with infection or inflammatory processes; it was measured by nephelometry (Array Protein System); the upper limit for defining a positive case was 0.8 mg/dL. Subjects with clinical manifestations suggestive of infection or inflammatory processes were excluded from blood sampling. The evaluation of iron status included measurements of hemoglobin and mean corpuscular volume (CELL-DYN 1700; Abbott Diagnostics, Abbott Park, IL), free erythrocyte protoporphyrin (ZP Hematofluorometer model 206D; AVIV Biomedical Inc, Lakewood, NJ), and serum ferritin (11). Subjects with anemia, defined as a hemoglobin concentration <120 g/L (women) or <130 g/L (men) or neutrophilia (>8 x 106/L), were also excluded. Liver function tests [ie, measurements of serum glutamic-oxaloacetic transaminase (EC 2.6.1.1), serum glutamic-pyruvic transaminase (EC 2.6.1.2), and serum -glutamyltransferase (EC 2.3.2.2)] were conducted by using commercial kits (Química Clínica Aplicada SA, Amposta, Spain).
Data generated during the 2-mo water consumption survey were used to calculate more precisely the subjects’ copper exposure. Assuming that copper ingestion from food was similar in all groups, the results of the blood studies were expressed per quartile distribution of copper ingested in water.
Analysis
The reports of gastrointestinal symptoms were analyzed by treatment group (<0.01, 2, 4, or 6 mg Cu/L), including all the participants (n = 1365). The results of the blood assessments were analyzed in 2 ways: by treatment and by quartile distribution of the variables measured (n = 195). Means ± SDs and quartile distributions were calculated and analysis of variance, chi-square analysis, and correlation analysis were conducted by using SAS software (SAS Institute Inc, Cary, NC).
RESULTS
Data recorded in the diaries showed that 16.3% of the subjects reported at least one symptom on one occasion during the 2-mo survey. On the basis of chi-square analysis (with Bonferroni correction) and a comparison of the 4 groups, the percentage of subjects reporting gastrointestinal symptoms was significantly greater in the group that consumed 6 mg Cu/L water than in the group that consumed <0.01 mg Cu/L (Table 1).
View this table:
TABLE 1 . Gastrointestinal symptoms in the 4 groups during the 2-mo controlled copper exposure
Of the 240 participants that provided a blood sample after the 2-mo copper-loading period, 45 had anemia, signs of infection or inflammatory processes, or both and were excluded from further analysis. No significant differences were observed in the proportion of excluded persons among the 4 study groups. The age, sex, and education level of the subjects were not significantly different between the 4 groups: the mean age was 32.9 ± 12.5 y, the mean duration of education was 8.9 ± 3.2 y, and 48.2% of the participants were men.
As expected, copper intakes from water differed between the study groups (Table 2). Analysis of variance showed no significant differences between the 4 groups in serum, erythrocyte, and mononuclear cell copper concentrations; in ceruloplasmin; in the nonceruloplasmin copper pool; in SOD activity; or in the indicators of liver function (Figures 1–3 and Table 2). The ranges of the different variables measured in the group that ingested 6 mg Cu/L were not significantly different from those of the group that ingested <0.01 mg Cu/L water. An analysis of the 195 participants individually showed no correlation between copper exposure and the laboratory measurements and no significant differences by sex or age. The nonceruloplasmin copper pool and serum copper concentrations showed a direct correlation (Pearson’s r = 0.58, P < 0.001). Copper concentrations in mononuclear cells did not correlate with serum copper or ceruloplasmin concentrations. The numbers of participants with values above or below the cutoff for the different variables evaluated are shown in Table 3. In 9.2% of the participants, liver enzyme activities were above the normal limits; however, the increased values observed were of no clinical relevance, and none of the participants had clinical evidence suggesting a liver disorder. In addition, in all cases, the out-of-range value was for only one of the enzymes measured. Measures of copper status and ceruloplasmin were not significantly different between the participants who reported symptoms and those who did not.
View this table:
TABLE 2 . Copper-load indicators and serum liver enzyme concentrations by quartile (Q) of copper intake in drinking water1
FIGURE 1. . Distribution of serum copper by quartiles (Q1–Q4) of copper intake in water in healthy adults (n = 49 per group). There were no significant differences between the quartiles.
FIGURE 2. . Distribution of serum ceruloplasmin by quartiles (Q1–Q4) of copper intake in water in healthy adults (n = 49 per group). There were no significant differences between the quartiles.
FIGURE 3. . Distribution of nonceruloplasmin copper by quartiles (Q1–Q4) of copper intake in water in healthy adults (n = 49 per group). There were no significant differences between the quartiles.
View this table:
TABLE 3 . Percentage of subjects with abnormal laboratory indicators by quartile (Q) of copper intake in drinking water1
DISCUSSION
Consumption of beverages or drinking water contaminated with copper has been associated with the occurrence of gastrointestinal upsets (12). Nausea, vomiting, diarrhea, and abdominal pain are commonly described as symptoms of toxic effects of acute copper exposure. Previous clinical trials showed that nausea is the most frequent and earliest symptom reported, showing a significant increase at 4 mg Cu/L (as copper sulfate) (7, 8). In the current study we tested a larger number of subjects. Although the same gastrointestinal symptoms were detected at all 4 copper exposures, the incidence of gastrointestinal symptoms was only significantly greater than baseline at 6 mg Cu/L water. For the sake of comparison, no correction was used in the comparison among groups in these 2 previous studies; if the Bonferroni correction had not been applied to the data generated in the current study, the incidence of symptoms would also have been significantly greater at 4 mg Cu/L water than at baseline (7, 8). When gastrointestinal symptoms in response to a subacute copper exposure were compared with blood measurements in the same subjects, there was no correlation between blood variables and reported symptoms.
In the past few years we have gained insight into the adverse effects of acute copper exposure (7, 8, 13), but the effects induced by longer (or more chronic) copper exposure remain unknown. In the current study, the exposures tested proved safe on the basis of measured functional liver enzymes. Coarse indicators of liver damage—such as serum glutamic-oxaloacetic transaminase, serum glutamic-pyruvic transaminase, and serum -glutamyltransferase—showed no changes in liver function. However, it may be possible that the changes were mild and thus undetectable. It is possible that intestinal absorption decreased or that biliary excretion increased in response to greater copper availability. It is also possible that copper deposits in hepatocytes increased but that the magnitude of this increase over the 2-mo study did not affect the large liver homeostatic capacity. It is crucial to define the early liver response to increased copper exposure. To what extent these early responses are reversible and what mechanisms are involved in the appearance of tissue damage should be addressed in the future to define chronic adverse effects.
Copper requirements to maintain physiologic copper metabolism have been controversial for a long time. In 1989 the National Research Council (United States) established an estimated safe and adequate dietary copper intake of 23.6–47.2 µmol/d (1.5–3 mg/d) for adults (14). However, studies carried out with Western diets showed that >30% of the diets tested by chemical analysis provided <15.7 µmol Cu/d (1 mg Cu/d) (15), the approximate amount of dietary copper that proved to be insufficient in some subacute (108–120 d) copper-depletion experiments in humans (2). The lack of apparent disorders in the population suggests that either the recommendation was overestimated or that current markers of copper status are too coarse to detect mild biologically significant changes. In the current study, copper intake in the control group was estimated at 14.2 µmol/d (0.9 mg/d) from the diet and at 0.2 µmol/d (0.015 mg/d) from water. No changes in dietary habits relevant to copper consumption were detected in the study area over the past several years; therefore, it can be assumed that this intake represents the long-term copper exposure of this population. The results of this study support the idea that the recommendation was indeed overestimated and that the new value of 14.2 µmol/d (0.9 mg/d) recently set for adults is more realistic (16).
Serum copper and ceruloplasmin concentrations and SOD activity are typically used to assess copper status. In our study, no relation was found between copper-status variables and copper intake. SOD activity would reflect metabolically active copper or copper stores and long-term status. Measurement of the activity of this enzyme in a healthy population has proved helpful in comparisons of groups with a low-copper status with those with a higher or normal copper status (17, 18). Of interest is that SOD activity did not change in that study, supporting the idea that there may be a threshold below which copper stores are depleted and plasma indicators change (19, 20).
Nonceruloplasmin copper has been proposed by some as a marker of copper excess (21). This is controversial; however, because it is the only marker of copper excess proposed, we think it deserves discussion. A high nonceruloplasmin copper fraction (10 µmol/L) has been reported in cases of Wilson disease and other forms of childhood cirrhosis by Eife et al (21). Subjects participating in this study were apparently healthy, and their nonceruloplasmin copper concentrations (1.8 µmol/L) were well below those described as "high" in children by Eife et al (21).
Several human cell lines tested have been shown to increase their copper content in response to variations in copper in the medium (22–25). In the current study, the hypothesis of which was that blood cells would respond in a similar fashion to copper availability in the plasma, the copper content was determined in 2 different cell types. However, the lack of detectable variations means that this approach was not correct and that both erythrocyte and mononuclear copper contents are not suitable indicators of copper load.
The lack of detectable changes in the indicators of copper status measured in the current study can be interpreted as evidence of an adaptive response to copper intake or as evidence that the biochemical indicators traditionally used to assess copper status are inappropriate within the range of copper exposure evaluated. The identification of appropriate biomarkers for early and mild-to-moderate excess copper exposure represents a major challenge for the research of human copper nutrition.
ACKNOWLEDGMENTS
The article was written primarily by MA, MO, and MG, with the assistance of the other authors. All authors approved the final version of the manuscript and contributed to the design and implementation of the study and to the analysis and discussion of the results. The authors had no financial or personal interest in the organizations that sponsored the research.
REFERENCES