Amalgam studies: disregarding basic principles of mercury toxicity.

Dental amalgam, which has been used for over 150 years in dental practice, consists of about 50% metallic mercury. Studies on animal and humans show that mercury is continuously released from dental amalgam and absorbed by several body tissues. It is widely accepted that the main source of mercury vapor is dental amalgam and it contributes substantially to mercury load in human body tissues. There is still a controversy about the consequences of this additional mercury exposure from amalgam to human health. Many studies were performed to evaluate possible adverse effects. In this comment, these studies were analyzed with regard to their methodical quality by considering the newest findings on mercury toxicity and metabolism. In sum, a number of studies are methodically flawed drawing inaccurate conclusions as to the safety of dental amalgam.

The Mt. Diwata study on the Philippines 2000-treatment of mercury intoxicated inhabitants of a gold mining area with DMPS (2,3-dimercapto-1-propane-sulfonic acid, Dimaval).

Ninty-five inhabitants of the gold mining area of Mt. Diwata (on Mindanao, Philippines), who were diagnosed to be mercury (Hg) intoxicated, were orally treated with 2 x 200 mg of the chelating agent DMPS (Dimaval, Co. Heyl, Germany) for 14 days in the course of a UNIDO project focusing on mercury pollution abatement. Blood and urine samples before and after treatment, urine after the first application of DMPS and a hair sample were collected and analyzed for Hg. Before and after treatment extensive anamnestic data were collected, medical and neurological investigations and some neuro-psychological tests were performed. In spite of the short time of treatment most of the patients reported a marked improvement of the complaints which were stated by them before the therapy and which are characteristic for a chronic Hg intoxication, for example tremor, loss of memory, sleeplessness, metallic taste, etc. But even in some of the objective neurological parameters like hypo-mimia, Romberg test and tests for tremor/ataxia a statistical significant improvement could be found. Significant improvements could also be found in two neuro-psychological tests (pencil tapping and Frostig). In some cases an extreme high urinary Hg excretion was found under the chelating therapy with DMPS, and by this a distinct reduction of the Hg body burden. Nevertheless, in most cases Hg in blood and urine was not markedly decreased by the treatment. This shows that the duration of the treatment (14 days) was not sufficient for a permanent decrease in Hg. As DMPS excretes Hg mainly through the kidney, it can be concluded that in most cases even after 14 days of treatment there was an ongoing redistribution of Hg from other tissues to the kidney. In conclusion, this study proves that a chelating therapy with DMPS is highly effective even in the case of a mixed chronic and acute intoxication with an unknown combination of Hg vapor, inorganic Hg and organic Hg=methylmercury (MeHg), as characteristic for gold mining areas in the third world. Adverse side effects were rarely reported. Only in one case the medication had to be terminated after the first application due to an allergic skin reaction.

The Mt. Diwata study on the Philippines 1999--assessing mercury intoxication of the population by small scale gold mining.

The region of Diwalwal, dominated by Mt. Diwata, is a gold rush area on Mindanao (Philippines) where approximately 15000 people live. The fertile plain of Monkayo is situated downstream, where people grow crops such as rice and bananas; locally caught fish is eaten frequently. The ore is dug in small-scale mines and ground to a powder by ball-mills while still in Diwalwal. The gold is then extracted by adding liquid mercury (Hg), forming gold-amalgam. To separate the gold from the Hg, in most cases the amalgam is simply heated in the open by blow-torches. A high external Hg burden of the local population must be assumed. To evaluate the internal Hg burden of the population and the extent of possible negative health effects, 323 volunteers from Mt. Diwalwal, Monkayo and a control group from Davao were examined by a questionnaire, neurological examination and neuro-psychological testing. Blood, urine and hair samples were taken from each participant and analyzed for total Hg. A statistical evaluation was possible for 102 workers (occupationally Hg burdened ball-millers and amalgam-smelters), 63 other inhabitants from Mt. Diwata ('only' exposed from the environment), 100 persons, living downstream in Monkayo, and 42 inhabitants of Davao (serving as controls). The large volume of data was reduced to yes/no decisions. Alcohol as a possible bias factor was excluded (level of alcohol consumption and type, see Section 4.4). Each factor with a statistically significant difference of at least one exposed group to the control group was included in a medical score (0-21 points). In each of the exposed groups this score was significantly worse than in the control group (median control, 3; downstream, 9; Mt. Diwata, non-occupational exposed, 6; Hg workers, 10). In comparison to the surprisingly high Hg concentration in blood (median, 9.0 microg/l; max, 31.3) and in hair (2.65 microg/g; max, 34.7) of the control group, only the workers show elevated levels: Hg-blood median 11.4, max 107.6; Hg-hair median 3.62, max 37.8. The Hg urine concentrations of the occupational exposed and non-exposed population on Mt. Diwata was significantly higher than in the control group: control median 1.7 microg/l, max 7.6; non-occupational burdened median 4.1, max 76.4; and workers median 11.0, max 294.2. The participants, living downstream on the plain of Monkayo show no statistically significant difference in Hg-blood, Hg-urine or Hg-hair in comparison with the control group. The German Human-Biological-Monitoring value II (HBM II) was exceeded in 19.5% (control), 26.0% (downstream), 19.4% (Mt. Diwata, non-occupational) and 55.4% (workers) of the cases, the German occupational threshold limit in 19.6% of the workers. Only some of the clinical data, characteristic for Hg intoxication (e.g. tremor, loss of memory, bluish discoloration of the gingiva, etc.), correlate with Hg in blood or urine, but not with Hg in hair. The medical score sum correlates only with Hg in urine. The poor correlation between the Hg concentration in the biomonitors to classic clinical signs of chronic Hg intoxication may be explained by several factors: Hg in blood, urine and hair do not adequately monitor the Hg burden of the target tissues, especially the brain. Inter-individual differences in the sensitiveness to Hg are extremely large. In this area a mixed burden of Hg species must be assumed (Hg vapor, inorganic Hg, methyl-Hg). Chronic Hg burden may have established damage months or even years before the actual determination of the Hg concentrations in the bio-monitors under quite different burden was performed (Drasch G. Mercury. In: Seiler HG, Sigel A, Sigel H, editors. Handbook on metals in clinical and analytical chemistry. New York: Marcel Dekker, 1994:479-494). Therefore, a 'Hg intoxication', that should be treated, was not diagnosed by the Hg concentration in the bio-monitors alone, but by a balanced combination of these Hg values and the medical score sum. In principle, this means the higher the Hg concentration in the bio-monitors, the lower the number of characteristic adverse effects are required for a positive diagnosis. By this method, 0% of the controls, 38% downstream, 27% from Mt. Diwata, non-occupational exposed and 71.6% of the workers were classified as Hg intoxicated. A reduction of the external Hg burden on Mt. Diwata is urgently recommended. An attempt to treat the intoxicated participants with the chelating agent dimercaptopropanesulfonic acid (DMPS) is planned.

Gold and palladium burden from dental restoration materials.

From 81 volunteers (16 without dental restorations, 65 with gold crowns or inlays) samples of saliva before and after chewing gum, blood, serum, urine and faeces were taken and analysed for gold (Au) and palladium (Pd). The Au concentration in all analysed biomonitors correlates significantly to the number of teeth with gold restorations. For Pd the correlations were still significant, but weaker than for Au. Persons with gold restorations show maximal Au and Pd concentrations, 10(2)-10(3) higher than the background burden. The calculated maximal daily Au load in saliva (1.38 mg Au per day) reaches the range of an oral Au therapy for rheumatoid arthritis with 6 mg Auranofin (= 1.74 mg Au per day). During this therapy severe and frequent side effects are reported. In contrast, the Au concentration in serum maximally reached from Au restorations, amounts to only approximately 1/20 of the Au level during arthritis therapy. But even under subtherapeutic doses of 1 mg Auranofin/day severe side effects have been reported (4 out of 56 cases). The mean Au blood concentration from 1 mg Auranofin daily was only 3 times higher than our maximum value. A toxicological classification of the Pd values is difficult, because no toxicological threshold limit has been established, especially for the low-level long-term burden with Pd.

Content of non-mercury-associated selenium in human tissues.

Recent studies have shown that at a higher mercury (Hg) burden, the molar ratio of selenium (Se) and Hg in tissues tends to approximate 1:1 by the formation of biologically largely inert adducts. From the toxicological standpoint, this trapping of free Hg is welcome. However, this binding of Se to Hg reduces the portion of Se in tissues, which is available for the formation of essential selenoenzymes like glutathione peroxidase, type I deiodase, and so forth and could result in a relative deficiency of Se. Therefore, we tried to determine the concentration of non-Hg-associated Se in several human tissues. As there is no proved trace method for the speciation of non-Hg-bound and Hg-bound Se in tissues, the total concentrations of Hg and Se were determined and the portion of non-Hg-associated Se was calculated by the difference of the molar concentrations of Se and Hg. For this investigation, the following tissues were obtained by autopsy from 133 adults: kidney cortex, thyroid gland, liver, spleen, cerebrum cortex, and pituitary gland. In no case was an occupational Hg burden known. The results confirm the assumption of a 1:1 association of Hg and Se in human tissues. The mean concentration of non-Hg-bound Se was calculated to 576 microg/kg in the kidney cortex, 363 microg/kg in the thyroid gland, 308 microg/kg in the liver, 205 microg/kg in the spleen, 111 microg/kg in the cerebrum cortex, and 545 microg/kg in the pituitary gland. In none of the cases under investigation in any tissue was the molar Se/Hg ratio below 1. This means that a total deficiency of non-Hg-bound Se could not be seen in this normal population, even at a higher Hg burden. Nevertheless, at a suboptimal Se supply like in Germany, any reduction of the part of Se, which is available for the formation of essential seleno-enzymes, should be avoided. Therefore, any additional Hg burden such as from dental amalgam should to be considered critically. The different distribution of Hg and Se in the body confirms that there is a controlled hierarchy in the Se supply of different organs, which tries to prevent a Se deficiency in organs with essential seleno-enzymes like the thyroid gland even under an suboptimal Se supply.

Mercury in human colostrum and early breast milk. Its dependence on dental amalgam and other factors.

The mercury concentration in 70 breast milk samples (Hg-M) from 46 mothers, collected within the first 7 days after delivery, was determined by cold vapour atomic absorption spectrometry. For comparison, 9 formula milk samples (reconstituted with Hg-free water) were investigated. The Hg-M in the human milk samples ranged from < 0.2 to 6.86 micrograms/L (median 0.37), in the formula milk samples from 0.4 to 2.5 micrograms/L (median 0.76). The Hg-M in the breast milk samples correlates positively with the number of maternal teeth with dental amalgam. The mean Hg-M of amalgam-free mothers was < 0.2 microgram/L, while milk from mothers with 1-4 amalgam fillings contained 0.57 microgram/L, with 5-7 fillings 0.50 microgram/L and with more than 7 fillings 2.11 micrograms/L. Hg-M correlated negatively to the day after delivery. Frequency of fish consumption tends to influence Hg-M positively, while the age of the mother shows no significant correlation. In the first 2 to 3 days after delivery some colostrum samples with Hg-M higher than in formula milk were found. Later on, the Hg-concentration in the breast milk was equal or even lower to that in formula milk. The higher Hg burden of infants' tissues from mothers with dental amalgam, as reported previously, must be explained (1) by a prenatal transfer of Hg from the mother's fillings through the placenta to the fetus, followed by a redistribution of this Hg in the body of the newborn, and (2) an additional burden via breast milk. Nevertheless, the comparison of Hg-M in breast and formula milk, the relatively moderate Hg burden in both kinds of milk, and the multiple manifest advantages of breast feeding speak against any limitation of nursing, even for mothers with a large number of dental amalgam fillings.

Increased blood mercury levels in patients with Alzheimer's disease.

Alzheimer's disease (AD) is a common neurodegenerative disorder that leads to dementia and death. In addition to several genetic parameters, various environmental factors may influence the risk of getting AD. In order to test whether blood levels of the heavy metal mercury are increased in AD, we measured blood mercury concentrations in AD patients (n = 33), and compared them to age-matched control patients with major depression (MD) (n = 45), as well as to an additional control group of patients with various non-psychiatric disorders (n = 65). Blood mercury levels were more than two-fold higher in AD patients as compared to both control groups (p = 0.0005, and p = 0.0000, respectively). In early onset AD patients (n = 13), blood mercury levels were almost three-fold higher as compared to controls (p = 0.0002, and p = 0.0000, respectively). These increases were unrelated to the patients' dental status. Linear regression analysis of blood mercury concentrations and CSF levels of amyloid beta-peptide (A beta) revealed a significant correlation of these measures in AD patients (n = 15, r = 0.7440, p = 0.0015, Pearson type of correlation). These results demonstrate elevated blood levels of mercury in AD, and they suggest that this increase of mercury levels is associated with high CSF levels of A beta, whereas tau levels were unrelated. Possible explanations of increased blood mercury levels in AD include yet unidentified environmental sources or release from brain tissue with the advance in neuronal death.

Correlation of mercury and selenium in the human kidney.

The total mercury (Hg) and selenium (Se) concentrations were determined in kidney cortex samples of 195 deceased, non-occupationally burdened individuals. Mercury was determined by means of Cold-vapour Atomic Absorption Spectrometry (CV-AAS) and selenium by Graphite-Furnace Atomic Absorption Spectrometry (GF-AAS). The molar Se/Hg ratio is high (up to (a) 300) in cases with relatively low mercury concentrations [Hg]. The ratio decreases with increasing [Hg]. At [Hg] of 700-1000 ng/g it reaches unity, where it remains constant even at larger [Hg]. Since in vitro mercury and selenium form relatively stable adducts, our results suggest the formation of a 1:1 Hg-Se compound that may explain mercury detoxification by selenium. This effect also results in the trapping of available selenium by mercury, too. Decreasing the reserve of free (i.e. not Hg-bound) selenium. The effect of this decrease of free selenium is under further investigation.

Carboxyhemoglobin blood concentrations in suicides by fire.

A total of 21 suicides by fire (16 males and 5 females) were investigated. In at least 18 of the 21 cases, flammable liquids were used as accelerants leading to severe and extensive burns. The determination of carboxyhemoglobin concentrations revealed comparatively low levels of between 3 and 30 per cent in suicides committed in the open whereas four out of five deceased found dead in gutted cars showed carbon monoxide saturations ranging from 34 to 87 per cent. In 18 cases (86%) soot was detectable in the airways even in cases with slightly elevated carboxyhemoglobin concentrations, indicating the importance of a careful examination of the airways at autopsy for the detection of vital signs.

Silver concentrations in human tissues. their dependence on dental amalgam and other factors.

Human tissue samples (liver, kidney cortex, 5 brain regions: grey matter of cerebrum, white matter of cerebrum, nucleus lentiformis, cerebellum, brain stem) from 173 decreased persons were analysed for silver (Ag) by GF-AAS (Graphite Furnace Atomic Absorption Spectrometry) and the results compared with the number of teeth with amalgam fillings and the concentration of inorganic mercury (Hg), which had been determined in the same tissue samples in a previous study. It was found that the mean Ag concentrations in liver and brain of adult females are approximately twice that of males. Moreover, the Ag concentrations, especially in the brain, depend possibly on age. To exclude these confounding factors as far as possible, the influence of dental amalgam and the correlation of Ag and Hg were evaluated only in a sub-group of 93 males, aged 11-50 years. In this sub-group statistically significant correlations were found between the number of teeth with dental amalgam and the Ag concentrations in the cerebral cortex and the liver. No such correlation was found for the kidney. Ag and inorg. Hg correlate well in this sub-group in the liver, but not in the cerebral cortex or the kidney. Individuals from this sub-group with (i) 0-2 and with (ii) more than 9 teeth with amalgam fillings show mean Ag concentrations (micrograms/kg in tissue wet weight, geom. mean) of 1.59 and 5.41 in the grey matter of cerebrum, 1.42 and 4.25 in the white matter of cerebrum, 1.53 and 4.89 in the nucleus lentiformis, 1.95 and 5.02 in the cerebellum, 1.05 and 3.27 in the brain stem, 3.40 and 8.15 in the liver and 0.42 and 0.44 in the kidney cortex. In contrast, comparing all individuals under investigation with only 0-2 teeth with amalgam no correlation between Ag and inorg. Hg could be found in liver, kidney cortex or cerebral cortex. These results show that amalgam fillings release Ag as well. Considering the different toxicokinetics of Ag and Hg it can be concluded that Ag is a reliable marker for the fact that the elevated concentrations of inorg. Hg found in tissues of individuals with amalgam fillings derive mainly from these fillings and not from other theoretically possible sources.

Mercury burden of human fetal and infant tissues.

UNLABELLED: The total mercury concentrations in the liver (Hg-L), the kidney cortex (Hg-K) and the cerebral cortex (Hg-C) of 108 children aged 1 day-5 years, and the Hg-K and Hg-L of 46 fetuses were determined. As far as possible, the mothers were interviewed and their dental status was recorded. The results were compared to mercury concentrations in the tissues of adults from the same geographical area. The Hg-K (n = 38) and Hg-L (n = 40) of fetuses and Hg-K (n = 35) and Hg-C (n = 35) of older infants (11-50 weeks of life) correlated significantly with the number of dental amalgam fillings of the mother. The toxicological relevance of the unexpected high Hg-K of older infants from mothers with higher numbers of dental amalgam fillings is discussed. CONCLUSION: Future discussion on the pros and cons of dental amalgam should not be limited to adults or children with their own amalgam fillings, but also include fetal exposure. The unrestricted application of amalgam for dental restorations in women before and during the child-bearing age should be reconsidered.

[Comparison of the body burden of the population of Leipzig and Munich with the heavy metals cadmium, lead and mercury--a study of human organ samples]. / Vergleich der Belastung der Bevölkerung im Leipziger und Münchner Raum mit den Schwermetallen Kadmium, Blei und Quecksilber--eine Untersuchung an menschlichem Organmaterial.

In the former GDR it was almost impossible to perform studies on environmental pollution. Therefore, a short time after the opening of the German interior border the authors started an investigation on the heavy metal burden of humans in the former District of Leipzig (Saxonia). In 1990/91 tissues from 57 deceased were collected from this region. The concentrations of cadmium were determined in specimens from the liver and renal cortex, of mercury in liver, renal cortex and grey matter of the cerebrum and of lead in samples of the pelvic bone and the cortical part of the femur. After sample pretreatment, the heavy metal concentrations were determined by GF-AAS or CV-AAS. The results were compared with studies recently performed by the authors in the region of Munich (southern Bavaria, FRG) and with values from the literature. It was found that the burden with cadmium was similar in both regions, whereas on the average, the bone lead concentrations in Leipzig were 10 times (!) higher than in Munich. Considering dental amalgam as main factor for the burdening with inorganic mercury, the mercury concentrations in the liver and the renal cortex were of the same order of magnitude in both regions. In contrast to this, significantly higher mercury concentrations were found in the brain samples from Leipzig than from Munich. Possible negative health effects of these elevated lead and mercury burden in the Leipzig area are discussed.

Drug death autopsies at the Munich Institute of Forensic Medicine (1981-1992).

A total of 638 drug death autopsy cases in southern Bavaria from 1981 to 1992 were analysed, including epidemiological and toxicological investigations. The rate of HIV infections decreased during the last few years. Cocaine does not (yet) play a major role. Suicide rates are high. Heroin intoxications are the most frequent cause of death, mostly in combination with other drugs and alcohol. In 1992 we observed a sharp increase of the number of deaths associated with dihydrocodeine abuse. This seems to be a local phenomenon and has to be explained by uncritical and uncontrolled prescription of large amounts of this opiate by individual physicians.

Clinical and morphological findings on mustard gas [bis(2-chloroethyl)sulfide] poisoning.

In 1984 and 1985, a total of eleven Iranian patients were transferred to hospitals in Munich, Germany, after a reported gas attack in the Iran-Iraq war. The initial symptoms and pretreatment in Teheran, Iran, as well as the admittance examination data, the clinical courses of the patients, and the clinical laboratory data in Germany, are reported. The main injuries were to the skin, the eyes, and the respiratory tract. One patient stopped breathing suddenly on the third day of treatment (eight days after the exposure). A large piece of mucous membrane blocking a bronchus was removed during an immediate bronchoscopy, but attempts at resuscitation failed. The most important autopsy findings in this case were severe pseudomembranous inflammation of the trachea and the bronchial tubes. The histological findings are reported. Chemical proof of the poison (mustard gas) was established. A review of the history of chemical warfare, the physical and chemical properties of mustard gas, and a literature survey of clinical findings (including, especially, experiences from World Wars I and II) contribute to the understanding of the actual cases.

Metallothionein, copper and zinc in fetal and neonatal human liver: changes during development.

Total and cytosolic zinc (Zn) and copper (Cu), cytosolic metallothionein (MT) and the Cu-load of MT were investigated in fetal (22, 24 and 32 gestational weeks) and neonatal (2-15 months) human liver. Whereas the fraction of cytosolic Zn remained constant at 66% of the total independent of the stage of development, the fraction of cytosolic Cu increased from 26% in preterm liver to about 100% within 12 months postnatally. The MT content was higher in fetal than in neonatal liver. There was a linear correlation (r = 0.996) between cytosolic MT and Zn in both fetal and neonatal liver but not between MT and Cu. In contrast to fetal liver, the Cu-load of MT in neonatal liver seems to be determined by the Zn/Cu ratio in the cytosol. The results suggest that MT is involved in the regulation of Cu and Zn metabolism during fetal and neonatal development.

Metallothionein in human liver and kidney: relationship to age, sex, diseases and tobacco and alcohol use.

The content of metallothionein (MT) was determined in the livers (MT-L) and the kidneys (MT-K) of 145 deceased persons. The individual values showed marked variations; average levels (arithmetic means +/- S.D.) were 154.9 +/- 151.4 mg/kg liver wet wt. and 160.5 +/- 150.4 mg/kg total kidney wet wt. In contrast to MT-L, MT-K increased with age up to a maximum around mid-life and decreased at higher ages. Neither the MT-L nor the MT-K depend on sex. The MT content significantly correlates with the macroscopic and histopathologic status of the liver. As compared to normal tissue, livers with fatty degeneration, cirrhosis and brown atrophy show MT-L values of 40%, 25% and 233% respectively. Low MT-L of 40% of the control values are also observed in individuals with a history of alcohol abuse. This decline preceded visible macroscopic pathological findings. MT-L also responds to the cause of death. At suicidal overdose of drugs the MT-L is lowered to 62% and at "mechanical" suicides it is increased to 137%. Whereas MT-K is essentially independent of the kidney status, it is significantly increased to approximately 190% in male smokers.