Maternal and neonatal hair mercury concentrations: the effect of dental amalgam

Authors


* S. W. Lindow, Hull Maternity Hospital, Hedon Road, Hull, HU9 5LX, UK.

Abstract

Objective To evaluate maternal and fetal hair mercury levels in relation to the placement of dental amalgam tooth restorations.

Design Cross sectional study involving women who never had dental amalgam restorations placed, women who had amalgam restorations placed before pregnancy and women who had restorations placed during the index pregnancy.

Setting North of England Maternity Hospital.

Sample Fifty-three healthy women who delivered healthy babies at term.

Methods Maternal and fetal hair was collected in a standardised manner in the first few days following delivery.

Main outcome measures Maternal and neonatal hair mercury concentrations.

Results When compared with women without restorations, there was a significant increase in the maternal hair mercury concentration in women who had dental amalgam placed outside of the index pregnancy and also in women who had dental amalgam placed during the index pregnancy. The fetal hair mercury concentration was significantly higher in babies when mothers had been exposed to dental amalgam either before pregnancy or during pregnancy compared with unexposed babies. There was no difference in the maternal or fetal hair mercury levels in the groups of patients who had dental amalgam placed before or during pregnancy.

Conclusion Maternal and fetal hair mercury levels were significantly higher in women who previously had dental amalgam restorations placed. There was no evidence that placement of dental amalgam restorations in pregnant women who had already similar restorations increased the maternal or fetal hair mercury level.

Introduction

The use of dental amalgam containing mercury is controversial and furthermore the use in pregnant women has been the subject of advice from the Chief Dental Officer ‘to avoid, where clinically reasonable, the placement or removal of amalgam fillings during pregnancy until appropriate data are available’1.

The British Dental Association Fact File2 states that although there is no evidence of any link between amalgam use and birth defects or stillbirths, it is sensible to minimise health interventions during pregnancy and breastfeeding where this is clinically feasible. Broadly similar advice is available from the Canadian Dental Association3. There is however no evidence of any side effect of dental amalgam4.

Previous work done in the area of East Yorkshire found that 64% of women underwent a dental check up during pregnancy and of those that went to a dentist 27% had dental restorations placed in their teeth. This amounts to 17% of the total of 496 women who took part in that survey5.

Mercury from dental amalgam restorations has been shown to cross the placenta in the sheep and accumulate in fetal tissues to achieve a steady state with advancing gestation6.

In human studies, dental amalgam ‘silver’ tooth restorations has been shown to increase blood mercury concentrations when compared with subjects without any dental amalgam restoration7. Additionally, it was also shown that the blood mercury level in those volunteers with amalgam restorations was positively correlated with the number and surface area of the restorations.

The transfer of mercury across the placenta in human subjects is somewhat more difficult to study than in animals. The concentration of mercury in maternal or fetal blood and amniotic fluid may not represent the chronic situation particularly if mercury amalgam restorations have been placed during the index pregnancy.

Scalp hair has been recognised as a valuable biopsy material for assessing mercury status and it is possible to use this biopsy material to evaluate the fetal mercury status by coupled cold vapour atomic absorption and inductively coupled plasma mass spectrometry8.

This study was undertaken to look at the chronic transfer of mercury from mother to fetus during pregnancy by evaluation of maternal and fetal hair specimens. The intention was to relate prior dental treatment with dental amalgam and dental treatment in pregnancy to the transfer of mercury across the placenta.

Methods

Permission was obtained to perform this study from the local ethics and research committee. All women gave written informed consent to participation in the study.

Maternal hair samples were all taken in an identical manner from the fine hairs at the nape of the neck. Steel scissors were used to cut the hair and it was marked with thread so that the cut ends were all identified. Fetal hair specimens were all taken in an identical manner by cutting the hair on the back of the baby's head with steel scissors. The specimens were all preserved in sterile plastic containers and were ultimately analysed in the laboratory.

All women who agreed to participate in the study underwent a dental examination with emphasis placed on the evaluation of the number of dental restorations using dental amalgam. In addition, the number of tooth surfaces restored using dental amalgam was recorded. Restorations placed in pregnancy were recorded in a similar manner.

A questionnaire was given to the participants which detailed the use of hair products and fish consumption.

Patients where categorised into three groups.

  • 1Women who have never had any dental restorations with dental amalgam.
  • 2Women who have had dental restorations placed in the past but no restorations placed during the index pregnancy.
  • 3Women who had dental restorations placed before the pregnancy and also more placed during the index pregnancy.

Statistical analysis was performed by an unpaired t test or Mann–Whitney U test as appropriate and a χ2 analysis or Fisher's exact test for categorical data.

Nitric acid was SpA grade from Romil Chemicals (Shepshed, Leicestershire, UK). The 1000 ppm mercury standard solution was Aristar grade (BDH Merck, Poole, Dorset, UK) and was used to generate the required standards. The diluent water was obtained from an Elga (High Wycombe, Buckinghamshire, UK) UHQ PS system and had a conductivity of 18 MΩ cm−1. All glassware was grade A, acid washed and soaked with 2% nitric acid. All solutions were immediately transferred into polypropylene vessels after making up to volume.

A CEM model MDS81D microwave oven with a separate pressure control line attached to one vessel was supplied by CEM Microwave Technology (Buckingham, UK). Samples were digested in 7 mL microvessels made from Teflon, placed inside Teflon PFA advanced composite vessels.

A thermoelemental (Winsford, Cheshire, UK) PQ2 + inductively coupled plasma mass spectrometry was used. The conventional nebulisation arrangement included a Cetac MCN 100 microcentric nebuliser with an uptake of 35 μl minute−1 and a quartz Scott type double pass spray chamber chilled to 4°C. The instrument was operated under the following conditions: nebuliser flow 0.88 L minute−1, cool gas 13.0 L minute−1, auxiliary flow 0.91 L minute−1, forward power 1350 W, reflected power <3 W. Mercury was analysed in triplicate at mass 200 and 202 amu in peak jumping mode, 3 points per peak, 10.24 ms per point for 30 seconds. The experimental limit of detection for mercury in solution, defined as 6× standard deviation of the blank analysis (three replicates) was 0.13 ppb (ng mL−1).

The samples of hair were prepared as follows. Approximately 20 mg of sample was accurately weighed (±1 μg) and placed in a 7 mL microvessel with 0.5 mL concentrated nitric acid. After an initial standing period of approximately 16 hours (overnight), during which time a loose cover was placed over the vessel to enable release of digest gases while preventing contamination, the top was sealed and it was placed in a 100 m advanced composite vessel. Two microvessels were placed inside each advanced composite vessel, which contained 10 mL water. A spacer at the bottom of the advanced composite vessel was used to avoid the microvessels being immersed in the water. The outer vessel was sealed and connected to the pressure control line of the microwave oven. The water in the outer vessel allowed the microvessel to reach a high pressure without venting by raising the gas pressure outside the microvessel. Up to six advanced composite vessels were added to the carousel for each digestion, but only one could be pressure controlled. The outer pressure control vessel was maintained at 100 psi for up to 30 minutes. After cooling to room temperature, the inner vessels were removed and further cooled in ice before being opened. The digest was diluted to 5 mL with water prior to analysis, incorporating 10 ppb, Ga, Rh, Ho and Bi as the internal standard.

Results

Sixty-three mother/infant pairs were initially taken for analysis. In 10 cases, contamination of the laboratory instrument rendered the results invalid and these mother/infant pairs were not included in the analysis. A further mother/infant pair is described separately below and therefore the data presented refer to 52 mother/infant pairs.

Three groups of women/infant pairs are described. Group 1 (n= 14) comprised women who had never had dental amalgam restoration. In group 2 (n= 29) were women who had dental amalgam restorations placed in the past but not during the current pregnancy. Group 3 (n= 9) comprised women who had dental amalgam restorations placed in the past and in addition further restorations had been placed during the index pregnancy.

The general characteristics of the groups were outlined in Table 1. It can be seen that women in group 1 were younger and of lower parity than women in the other two groups. This finding probably represents the increasing likelihood of dental caries and restorative treatment as women get older.

Table 1.  General characteristics of total study population. Values are expressed as n (%), mean [SD] or median {range}.
 Group 1 (n= 14)Group 2 (n= 29)Group 3 (n= 9)
  1. * One result missing.

  2. 1P < .05 (unpaired t test).

  3. 2P < .01 (unpaired t test).

  4. 3P < .01 (Mann–Whitney U test).

  5. 4P < .05 (Mann–Whitney U test).

  6. 5P < .05 (Mann–Whitney U test).

Age (years)22.6 [5.1]1,227.1 [4.3]228.9 [7.6]1
Height (cm)164.6 [6.7]165.4 [6.8]160.9 [5.3]
Weight (kg)64.4 [10.1]71.3 [16.4]*69.5 [16.6]
Gravidity1.0 {1–4}32.0 {1–6}31.0 {1–4}
Parity1.0 {1–3}42.0 {1–5}4,51.0 {1–2}5
Gestational age38.9 [2.0]39.8 [1.6]39.6 [1.7]
Smoking6 (42.9)8 (27.6)0
No. who eat seafood12 (85.7)29 (100)7 (77.8)
No. who eat seafood >1/week1 (7.1)10 (34.5)3 (33.3)

All three groups demonstrate a low consumption of seafood, although 85.7%, 100% and 77.8%, respectively, of women in the three groups ate seafood, only 14/52 ate seafood more then once per week. There was no significant difference in the seafood consumption between groups.

The dental history of the three groups of women is represented in Table 2. It can be seen that group 3 women had significantly more dental amalgam restorations (t=−3.7, P < .001) and more surfaces (t=−4.0, P < .0001) filled than the women in group 2.

Table 2.  Dental history of study groups. Values are expressed as mean (SD) or median [range].
 Group 1 (n= 14)Group 2 (n= 29)Group 3 (n= 9)
  1. 1P < .001 (unpaired t test).

  2. 2P < .0001 (unpaired t test).

Total no. of restorations4.9 (2.3)19.2 (4.9)1
Total no. of surfaces7.5 (4.0)216.3 (9.8)2
No. of restorations placed in index pregnancy1.7 (1.3)
No. of surfaces placed in index pregnancy3.4 (2.8)
Gestational age when restorations placed7.0 {4–30}

Table 3 outlines the hair mercury results. Group 1 had significantly lower maternal hair mercury levels than the women in group 2 (Mann–Whitney U test 65.5, P < 0.001, 95% CI −0.57 to −0.17). The women in group 1 also had significantly lower maternal hair mercury levels than women in group 3 (Mann–Whitney U test 26.0, P < 0.05, 95% CI −0.43 to −0.03). There was no significant difference between the maternal hair mercury level in the women in group 2 and group 3.

Table 3.  Hair mercury results. Values are expressed as n (%), mean [SD] or mean (median) [range].
 Group 1 (n= 14)Group 2 (n= 29)Group 3 (n= 9)
  1. *One case had both maternal and fetal levels of mercury undetectable and was therefore not included.

  2. 1P < 0.05.

  3. 2P < 0.001.

  4. 3P < 0.05.

  5. 4P < 0.05.

Maternal hair mercury (μg/g)0.19 [0.13]1,20.56 [0.51]20.43 [0.25]1
No. with undetectable maternal hair mercury1 (7.14)00
Fetal hair mercury (μg/g)0.17 [0.24]3,40.44 [0.45]40.45 [0.44]3
No. with undetectable fetal hair mercury5 (35.7)4 (13.8)0
Maternal/fetal mercury ratio1.36 (1.48) [0.31–2.15] (n= 9)1.54 (1.20) [0.28–3.94] (n= 25)1.51 (1.18) [0.24–2.96] (n= 9)
No. with maternal/fetal ratio >110 (77)*19 (66)5 (56)

There was a positive correlation between the number of fillings and the maternal hair mercury level (Spearman's rho 0.42, P < 0.01) and also between the number of surfaces and the maternal hair mercury level (Spearman's rho 0.37, P < 0.01).

The mean fetal hair mercury level was significantly lower in group 1 than in group 2 (Mann–Whitney U test 113, P < 0.05, 95% CI −0.48 to −0.06). Group 1 also had significantly lower fetal hair mercury levels than the women in group 3 (Mann–Whitney U test 29, P < 0.05, 95% CI −0.62 to 0.08). There was no significant difference between the fetal hair mercury levels in groups 2 and 3.

There was a positive correlation between the number of fillings and the fetal hair mercury level (Spearman's rho 0.33, P < 0.05), but not the number of surfaces and the fetal hair mercury levels (Spearman's rho 0.26, P= NS).

When all 53 patients are analysed together there is a positive correlation between maternal and fetal hair mercury levels (Spearman's rho 0.68, P < 0.01).

There were no differences in the maternal/fetal ratio among the three groups. The numbers in the maternal/fetal hair ratio group are correspondingly smaller as five babies and one women in group 1 had mercury levels below the limit of detection and four babies in group 2 also had undetectable levels. These cases did not have maternal/fetal ratios calculated but were assumed to have maternal/fetal ratios greater than 1. The percentages of cases with maternal/fetal ratios greater than 1 were 77%, 66% and 56%, respectively, in the three groups.

The one outstanding patient who was not included in either of the three groups had no dental restorations at all until she became pregnant in the index pregnancy when she had dental amalgam fillings placed. This woman had a maternal hair mercury level of 0.44 μg/g, a fetal level of 0.32 μg/g and a maternal/fetal ratio of 1.38. These three values lie within the ranges of the total sample of women.

Table 4 demonstrates the use of hair products (e.g. anti-dandruff shampoo or artificial colouring) and whether there were any hair treatments used in the last six months. There were no significant differences between the groups (Fisher's exact test).

Table 4.  Hair treatments. Values are expressed as n (%).
 Group 1 (n= 14)Group 2 (n= 29)Group 3 (n= 9)
  1. * Hair treatments included permanent wave, bleaching, straightening, anti lice treatments and artificial colour.

No. who use anti-dandruff shampoo1 (7.1)4 (13.8)1 (11.1)
No. who use artificial hair colourants6 (42.9)16 (55.2)4 (44.4)
No. who used any hair treatment* in the last six months5 (38.5) (n= 13)13 (48.2) (n= 27)2 (22.2) (n= 9)

In summary, women who never had dental amalgam restorations placed had significantly lower maternal and fetal hair mercury levels than women who previously had dental amalgam restorations. No difference could be detected in women or their baby's hair mercury levels when additional amalgam restorations were placed during the index pregnancy.

There were no significant differences in the maternal/fetal ratio of hair mercury concentrations in any of the three groups.

Discussion

Using maternal and fetal hair mercury concentrations, it has been shown that women who never had dental amalgam restorations have lower levels of mercury in their hair and also significantly lower levels of mercury in their baby's hair than women with dental amalgam restorations. There is no evidence that the placement of dental restorations during pregnancy added to either the maternal or fetal hair mercury levels.

In humans, mercury accumulation is from exposure to dental amalgam or via ingestion. Dental amalgam restorations lead to the inhalation of mercury vapour, which is oxidised to its inorganic form Hg2+ and appears in the blood. Blood mercury concentrations are correlated with the number and surface area of amalgam restorations and are lower in people without dental amalgam7. Inorganic mercury is uncommon in food but methyl mercury is found in seafood especially predatory fish and marine mammals and this represents the almost exclusive source in humans9.

This study, where total mercury was measured in hair samples, represents a situation where there is low fish consumption and dental sources represent a greater proportion of the total exposure.

The range of mercury in the maternal hair specimens was similar to the range described by Chai et al.10 in China. However, their mean value of 0.81 μg/g was almost double the present study mean. The present study demonstrated maternal hair mercury values much less than in hair samples collected from people around the Balbrina Reservoir in the Brazilian Amazon (mean 8.76 μg/g, range 2–21.6 μg/g)11. Maternal hair methyl mercury concentrations in 711 Seychellios mothers ranged from 0.5 to 26.7 μg/g (mean 6.8 μg/g)12, which was similarly much higher than the present cohort but both the Seychellese and Brazilian populations are large fish eaters and the Chinese population was from a mercury polluted area.

Maternal hair mercury levels have been found to correlate with fetal brain levels when six areas of the brain were studied in autopsy specimens13.

In a study of pregnant ewes, Vimy et al.6 showed that when dental amalgam is placed, there is evidence of mercury in maternal blood, fetal blood and amniotic fluid within two days. After 29 days, however, the levels of mercury plateau for the first 140 days after amalgam placement. As most of our study group had their dental work early in pregnancy, it is possible that the fetal hair, which is more abundant in later gestation, did not reflect this early exposure.

Adult tooth restorations funded by the National Health Service in England and Wales in 1995/1996 utilised dental amalgam in approximately 61.5% of cases (Dental Practice Board, personal communication). It is also notable that in the East Yorkshire survey in 1997, 17% of pregnant woman had dental restorations placed during pregnancy5.

There is such widespread use of dental amalgam the United Kingdom and if East Yorkshire is typical of the rest of the country, the placement of large numbers of restorations in the mouths of women who are pregnant raises the issue of safety.

Kuntz et al.14 demonstrated that a history of previous stillbirths and birth defects in pregnant women show a positive correlation with background mercury levels in maternal and umbilical cord blood specimens taken during pregnancy and postpartum. A contrary view is taken by Brodsky et al.15 who did not find any excess in the rates of miscarriage or congenital abnormalities in dental assistants who would be exposed to mercury occupationally when compared with the wives of dentists.

Female dentists were studied using a questionnaire and those in clinical practice had a higher miscarriage rate than controls derived from the class 1 group of the 1970 British Births Survey (12.8% vs 10.9%)16.

There is debate over the long term neurodevelopmental outcome of children who were exposed to high levels of mercury in utero. In a New Zealand cohort followed up to the ages of six to seven years, there was no correlation between high prenatal mercury exposure and educational performance on 26 scholastic and psychometric tests17. Similarly in the Seychelles, children studied at 66 months of age showed no adverse effects of prenatal mercury exposure from fish12.

Evidence of harm comes from studies relating to acute high level exposure. In Iraq, children were exposed in utero when their mothers consumed methyl mercury treated seed grain which was made into bread. It was suggested that almost all of the neurological findings were associated with mercury levels in maternal hair of 99–384 ppm (μg/g)18. These levels are approximately 200 times more than the maternal levels found in the current study.

In summary, using maternal and fetal hair analysis, a British urban cohort of women demonstrated low levels of hair mercury concentrations. The maternal and fetal hair mercury levels were significantly higher in women who previously had dental amalgam restorations placed. There was no evidence that the placement of dental amalgam restorations in pregnant women who already had similar restorations increased the fetal or maternal hair mercury level.

Ancillary