HEALTH EFFECTS OF DIOXINS

Everyone in industrialized countries has a potent mixture of dioxins, furans, co-planar PCBs, PCNs and other similar compounds stored and accumulated in their bodyfat. This chemical concoction of compounds in our bodies is likely to add together, making up a total dioxin-like toxicity: dioxins plus PCBs is equivalent to more dioxins.

In addition to these dioxin and dioxin-like molecules, we must also be concerned with other organochlorine compounds in our bodies which are not part of this family but are likely to interact with it. We do not know enough at present about these compounds which include non dioxin-like PCBs and pesticides such as DDT.

Everything that is said about dioxin also applies to the other members of the dioxin family of compounds - their differences lie only in the individual strength of their actions.

The most toxic member of the dioxin family is 2,3,7,8-TCDD and it is mostly mfrom studies on this compound that we know about the mechanism of the other chemicals in the same group (Silbergeld & Gasiewicz, 1989).

2,3,7,8-TCDD (often known simply as TCDD) is known for its lethal effects at mvery low concentrations: a millionth of a gram will kill a guinea pig. However, the reasons for its potency are very subtle, and connected with its structural similarity to potent natural hormones. The power of hormones lies in their ability to act in trace amounts as chemical messengers controlling vital processes in the body. Thus, an accidentally produced contaminant, i.e. TCDD, from the chemical industry can act as a wrong key in the subtle system of trace chemical messengers in the body by mimicking the action of a hormone.

TCDD produces a range of toxic effects:

* Lethal effects: animals die from a wasting disease in two to six weeks at levels ranging from 1 ug/kg bw to 5000 ug/kg bw * Immune system damage at similar levels in all animals examined, because of damage to the thymus gland causing changes in cell immunity: especially likely in children * Damage to other organs such as liver, kidney and digestive tract * Reproductive effects: miscarriage, sterility * Birth defects, including neurological effects * Cancer: most potent cancer promoter known, also evidence of some tumour initiation; animal carcinogen * Chloracne - persistent skin eruptions in humans and some animals

(Goldstein & Safe, 1989; Silbergeld & Mattison, 1987; Silbergeld, 1989; DOE, 1989)

Chloracne is a skin disease, often accompanied by severe disfiguration, severe joint pain, headaches, fatigue, irritability and chronic weakness; and it can persist in the body for at least 30 years after exposure (Kimbrough & Grandjean, 1989). No-one disputes that it is caused by dioxin-like compounds, but it is not an infallible marker of dioxin exposure (Gough, 1991).

In order to appreciate why the presence of dioxins in our bodies is so alarming, it is necessary to understand something about the mechanisms of cell activity with which it interferes. This chapter will first describe these, then go on to explain how dioxins act, with special reference to their effects before birth.

Action Of Dioxin-like Chemicals

The Messenger System - Receptors

Receptors are small masses, often made of protein, which exist within the cells in our bodies. Their function is to bind to molecules, forming a unit like a key in a lock. This unit can then influence the cell nucleus, or "brain" of the cell, often by becoming stuck to the DNA in the nucleus and controlling gene activity. The DNA then organises the production of useful chemicals such as enzymes or hormones, which go on to influence life maintaining chemical reactions in the cell. The importance of chemical messengers such as hormones is well known: these receptors are their essential partners.

The Ah Receptor

There are probably thousands of receptors: each has specific substances it has evolved to pair up with, although the 'partners' of many are not known. One of these receptors is the Ah, or aryl hydrocarbon, receptor. Its natural partner in the body has not been found, although similar receptors have been found in various cells in many animal species. This suggests that it may be an evolutionary relic: its partner may have been a growth regulating hormone that no longer exists (Silbergeld & Gasiewicz, 1989).

The present day function of the Ah receptor is to bind to many naturally occurring contaminants, e.g. benzene and other aromatic hydrocarbons, some plant toxins and all organochlorine compounds. The complex of receptor and contaminant then moves to the DNA in the cell nucleus where genes are activated to produce enzymes which break down the unwanted molecules. A 'family' of enzymes is produced, known as cytochrome P-450. One of the enzymes is AHH (aryl hydrocarbon hydroxylase) which catalyses a chemical reaction to initiate the breakdown of benzene and similar molecules. Once the molecule is degraded, enzyme production is switched off; the molecule is no longer on the receptor to give the signal. The Ah receptor thus plays an important part in detoxifying the body.

Much of this activity happens in the liver, although P-450 enzymes also occur outside this organ. Eventually, the "foreign" substances are converted to water soluble compounds that can be excreted by the kidneys.

The Key and Lock: The Receptor-Dioxin Complex

Many different molecules bind to the Ah receptor, but 2,3,7,8-TCDD binds extraordinarily well, as if it were the best possible key for the Ah receptor lock. It forms a strong unit with the receptor and this unit is known as the receptor-dioxin complex. Dioxins cannot act without the Ah receptor: all of the toxic actions depend on the formation of this complex and thus TCDD has no effect on animal cells which have no Ah receptors (Silbergeld & Gasiewicz, 1989). 2,3,7,8-TCDD is a flat molecule which resembles many natural hormones; perhaps it mimics the original hormone partner of the Ah receptor.

The relationship between dioxins and the Ah receptor is illustrated in figure 3.

Once the dioxin-receptor complex is formed, it moves to the DNA in the cell where it has two main effects:

1) It activates genes which produce P450 enzymes 2) It activates genes which produce substances that regulate the growth and division of cells. In this way dioxin is behaving like a hormone.

Dioxin and Enzyme Production

When a dioxin molecule is bound firmly to the Ah receptor and the receptor-dioxin complex moves to the DNA, the P-450 enzymes are stimulated. The resulting enzymatic activity includes attempts to break down the DNA-attached dioxin. If the dioxin is not successfully broken down, the receptor remains firmly stuck to the DNA and the P-450 enzymes are produced continuously, a process known as enzyme induction. This persistent activity has been seen in cases of occupational poisoning where abnormal liver enzymes have been found in workers exposed to TCDD over 20 years previously and where elevated AHH activity in the placentas of pregnant women poisoned by PCB-contaminated rice-oil 4 to 5 years earlier (Goldstein, 1989) has been reported. The toxicity of the different dioxin-like compounds mentioned above is dependent on the strength of their bond with the Ah receptor; 2,3,7,8-TCDD is the dioxin that binds the most strongly.

This stream of extra enzymes does not break down the dioxin appreciably, but it does produce the first stages in the breakdown of other chemicals, including hormones and the degradable chemicals the Ah system is good at dealing with. In many cases these intermediates are mutagens. So the effect of dioxin is to make other molecules carcinogenic, which at least partly explains its role as a cancer promoter. 2,3,7,8-TCDD is the most potent cancer promoter known (Silbergeld & Gasiewicz, 1989). TCDD makes the liver break down hormones more quickly and contraceptive failure has been attributed to this (Silbergeld & Gasiewicz 1989).

The Hormone-like Action of Dioxins.

The Ah receptor has functions other than enzyme induction. It also controls mgenes and cell functions, especially those regulating growth and the differentiation of cells. Dioxin is similar to natural hormone molecules such as thyroxin which is produced by the thyroid gland and which regulates the general metabolism of the body; both are flat aromatic molecules.

Thyroxin is unusual because it contains a carbon-iodine bond and is therefore one of the few natural molecules in the body to contain a halogen.

Natural hormones affect each other: they do not act in a vacuum; so dioxin's hormone-like activity may also indirectly alter the levels and activity of hormones not immediately acted on by dioxin. For instance, oestrogen needs to bind to a receptor to be active and TCDD has been found to decrease the number of oestrogen receptors in human cells (Safe 1980), thus reducing the effects of oestrogen (Safe et al.,1990). Certain oestrogen dependent cancers in rats were actually reduced by exposure to TCDD (Kociba, 1978).

Hormones, like dioxins, act in minute quantities: women's non-pregnancy oestrogen levels range from 61 to 437 parts per trillion of blood serum (Tietz, 1983). We have around 30 ppt of dioxin TEQs in our fat: this is around 1 ppt in blood serum.

The variety of effects seen in a wide range of organs in the body can be explained by the interference of dioxin with growth regulation. Growth occurs by means of cell division. Dioxins appear to affect the process by mwhich cells acquire individual characteristics. In the foetus especially the interference with this process whereby cells are organised can be devastating: resulting in birth defects such as cleft palate. They also cause uncontrolled growth of cells, something that suggests a mechanism for cancer promotion.

The dioxin-related effects which have been observed are all on epithelial cells, which form the lining of many different organs and parts of the body. The damage done by TCDD to the thymus gland, urinary tract, liver and bile ducts is caused by disruption of the growth of epithelial cells in these organs. The skin is made up of epithelial cells, and chloracne is caused by overgrowth and altered differentiation of skin cells. It is interesting to note that in the embryo, there is a close relationship between nerve cells (brain, spinal cord and nerves) and epithelial cells which form skin and organ linings. All these cells are formed from the same section of the early embryo.

Another mechanism that has been proposed for TCDD, that could help explain the birth defects seen after exposure, is that TCDD, by binding to the Ah receptor for too long, may produce a signal at the wrong time during organ formation. As a result, cells may become specialised too late, thus causing malformation (Silbergeld & Gasiewicz, 1989). Although dioxins have been shown to act exclusively through the Ah receptor, knowledge is not complete, and it is possible that there are further undiscovered mechanisms, such as action through another receptor.

The variation of TCDD effects between species and strains

The toxicity of TCDD in any species is dependent both on how many Ah receptors the animal in question has and also on the genes activated when the TCDD and Ah receptor are attached to the DNA. The response of DNA to the altered receptor also varies and is hereditary; with differences in susceptibility to TCDD also existing between strains (Byard, 1987).

The TCDD dose that produces rapid death varies a great deal between species: of the animals tested, the hamster is least susceptible, with a fatal dose of 5 mg /kg bodyweight while a dose 100 times smaller than this will kill a rat (Silbergeld & Gasiewicz, 1989).

Sub-lethal effects are seen at far lower doses, e.g. in mice, a dose which is 4000 times less than a fatal dose will significantly alter the ability of mice to produce cell-killing T-lymphocytes (a white blood cell important for immunity) (Silbergeld & Gasiewicz, 1989). Immune system effects are similar for various animals. TCDD induced changes in the thymus and liver occur at the same dose levels in the hamster as in otherwise more sensitive species (Silbergeld & Gasiewicz, 1989). The rat and the hamster are also very similar where birth defects are concerned. In both animals, a maternal dose of 1.5 ug/kg bodyweight produced different, but equally serious, effects in the foetuses, causing digestive tract bleeding in the rat and kidney abnormalities in the hamster (Olson, 1990).

How the toxicity of different dioxin-like compounds is compared

The toxicity of the various dioxin-like compounds depends on the strength of the binding to the Ah receptor (Silbergeld & Gasiewicz, 1989). This can be quantified by measuring the amounts of two of the enzymes produced when the receptor-dioxin complex is stuck to the DNA. These amounts are then compared to the amounts produced by 2,3,7,8-TCDD. With 2,3,7,8-TCDD assigned a value of one, the quantities of enzyme produced by each dioxin-like compound are then used to produce toxicity factors, the TEFs mentioned earlier.

This measure of toxicity, however, cannot predict the exact effects of each dioxin-like compound on cell division (and thus on birth defects), because these effects depend on the behaviour of genes and this varies between different cells and different species. It does not account for variations caused by the presence or absence of Ah receptors or for individual differences of ability to store pollutants in fat or to excrete them. Humans are especially diverse genetically. It is a guide only to the strength of the cell's reaction to dioxin and thus cannot accurately predict the impact of the dioxin on the body as a whole. As such, the TEF is an indicator of the wide ranging effects of dioxins on cell growth and differentiation: the more strongly the molecule binds to the receptor, the greater the chance of interference with scarcely known, delicate processes.

What are the combined effects of the dioxin family?

Unless it can be shown otherwise we must assume that dioxins, furans, PCBs and others will add together to give a total dioxin-like toxicity. This has been shown experimentally for dioxins and furans (Eadon, 1986), but doubts have been expressed about the validity of adding the PCBs' contribution in the same way (Goldstein & Safe, 1989). Studies have been carried out that show PCBs acting against and reducing TCDD effects. However, a moderately toxic PCB, which produced only mild pre-birth effects in mice on its own, was found to increase TCDD effects in those mice tenfold (Birnbaum et al., 1985).

In addition, PCBs are a mixture of dioxin-like and non-dioxin-like molecules. Some bind strongly to the Ah receptor and others only weakly. PCBs have been reported to exert other effects separately from their action through the Ah receptor: for example there are effects on nerve signal transmission linked to a PCB which binds only weakly to the Ah receptor (Seegal et al., 1990).

Certain PCBs can increase the levels of the Ah receptor in the liver, and hence increase susceptibility to low doses of other dioxin-like compounds(Goldstein & Safe, 1989). However, laboratory investigations of interactions such as these cannot possibly cover all that are possible in humans and animals exposed in the environment to the whole range of dioxin-like compounds. Considering both the evidence for enhancement and lessening of effects, we have to assume that the different compounds add together to produce a total dioxin-like toxicity, which can be estimated by using the TEFs given earlier. In other words, dioxins plus PCBs equals more dioxins.

Effects on Humans

It has been claimed that humans are far less susceptible to the effects of dioxin than other animals. Debate on this point, especially in the USA, has tended to distract from real assessment of the reasons for the apparent differences.

The effects of dioxins on human populations can not be examined using the same methods as animal experiments, for reasons that will be discussed. This absence of proof has led to assertions that "dioxins never killed anyone"; "dioxins cause only a skin rash in human beings". Humans are especially variable genetically, and therefore their sensitivity is likely to vary considerably.

Unlike small animals, humans do not appear to be susceptible to the rapidly fatal effects of dioxin. However, even here there is debate, because in a real human being it is always possible to assert that there must have been a pre-existing fatal disease. When there is one relatively rapid fatality after an accident, such as after Seveso (Cook, 1982) it will probably never be known whether it was caused by dioxin.

In the USA, a very low minimum risk intake was set for 2,3,7,8-TCDD, on the basis of cancer incidence in a large experiment on rats (Kociba, 1978). This limit has remained at 0.006 pg/kg bw/day until the present. It was based on the assumption that dioxin caused cancer in the same way as radioactivity: i.e. just one molecule on the DNA could have an effect, leaving no safe level, only a level which can be predicted to cause a certain risk of cancer. The US EPA has a standard risk, deemed tolerable, of one in a million over a 70 year lifetime, and this was used to set the limit. Since most people take in far more than 0.006 pg/kg bw/day, enforcement of this limit would have huge implications for the industries producing dioxins. So scientists have been employed to produce arguments against the EPA. A committee re-examined the Kociba slides and reclassified some of the tumours as benign. Millions of dollars were poured into investigations into the mechanism of dioxin. These have increased our knowledge but they have not established definitely whether or not dioxin causes severe health problems in human beings.

One of the results of these investigations has been gradual acceptance of the receptor mechanism for dioxin. This discovery has made us aware of the scarcely understood subtlety of the system in which very low levels of dioxin have their effects. However, a receptor-based mechanism is generally agreed to imply that a certain percentage of receptors must be occupied before effects can occur (Roberts, 1991). This has led to renewed attempts to claim that dioxins are not as toxic as was thought, fuelled by the desire of industry to declare present levels tolerable. The EPA has now begun to reassess its tolerable level, to take into account the likelihood of a threshold effect and it is possible that it will declare a 'safe' level. However, this will not be able to take into account all the contributions from dioxin-like compounds which individuals carry in their bodies; it still means that a precautionary view requires a reduction in our intakes and body levels. Scientists such as Ellen Silbergeld feel that some of those advocating a threshold approach are ignoring the complexities of dioxin's action. The old level may still be correct, even though it was set for the wrong reasons (Roberts, 1991). In order to assess health effects in humans we have to look at all the factors that are uncertain in real situations.

Although the receptor mechanism is widely accepted amongst scientists, there may be other pathways by which dioxin exerts its toxic effects, casting further doubt on the idea that there really is a threshold level.

It has been mentioned that immune system effects are much less variable between species than rapidly fatal effects. The effects on thymus cells, which play an important part in the immune response, occur at the same dose level in humans as in rodents (Silbergeld & Gasiewicz, 1989).

The role of dioxin in immune depression and reproductive effects may be difficult to prove in humans, because such conditions have multiple causes. For example, an increased incidence of infections may be recorded as being due to epidemics. Thus a gradual chemical-related decline in the health of the population may pass unnoticed.

In addition, everyone in industrialised countries has some exposure to the dioxin family, so it is difficult to define "exposed" and unexposed groups. It would be better to relate effects to actual exposure levels, but environmental and human levels have rarely been measured at the time of accidents.

Cancer has been the main concern with respect to dioxins. However, it takes 20 years or more for effects such as cancers to appear. A workforce suffering effects in 1991 could possibly be assessed for levels of dioxins now, but 20 years ago, the amount could not be measured with any degree of accuracy. The original amounts can be estimated using what is known about half-lives, but not for a whole collection of compounds. In spite of this, a study of workers exposed to dioxin in the U.S. has found increased mortality which is "consistent with a carcinogenic effect for TCDD" (Fingerhut et al,1991).

People do not form a uniform population: a control group which is relatively unexposed would have to match the exposed group in age, sex, socioeconomic status, pre-existing state of health and other characteristics. This contrasts with laboratory animals which can be quite easily standardised.

The full range of dioxin effects has not been looked for in human populations. We know that chloracne is one of the more rapid effects - it occurs within a few months of exposure. Most of the other effects which are known from animal experiments take a long time to show up in humans (eg. cancer) or they are known to have multiple causes (for example birth defects, and neurological disorders).

In addition to these difficulties, some studies of accidents are thought to have been biased by including unexposed workers in statistics (Rohleder, 1989) and possibly by intentional mis-design of surveys of ill-health (Jenkins, 1991).

Animal data indicates that the action of TCDD might be most critical for the developing immune system, especially just before and just after birth (DOE, 1989). Immune system effects were seen in children exposed to PCDF-contaminated PCBs through their mother's consumption of contaminated cooking oil (Rogan et al., 1988).

The easiest connections to prove between a toxic compound and a medical condition are those in which a poison produces a highly specific effect: a tumour, for example. TCDD produces a wide variety of different effects at different sites in the body, some of which may be too low to be statistically significant. The traditions of establishing causes, therefore, may be unsuited to finding the diverse effects from a poison like TCDD (Silbergeld & Gasiewicz, 1989). The possibility that effects are going unnoticed is supported by a significant increase in total cancer deaths in workers in dioxin-contaminated industries (Fingerhut et al., 1991), where certain individual cancers were not increased as expected.

All the evidence for the mechanism of TCDD's toxicity described above has been demonstrated in human cells or cell lines, so it is unlikely that humans are the only animals with low sensitivity to dioxins. Effects observed on human epithelial cells occurred at similar levels of TCDD to those which caused the same effects in animal cells.

Human tissue, ie. cells rather than the whole body system, seems to have a similar response to dioxins compared with other animals usually regarded as more sensitive, such as rats. One experiment, which compared the AHH enzyme induction in human placentas (from the Yucheng disaster) with that in rat livers, found that the human tissue was more sensitive: when the total toxicity from furans and PCBs in the placentas was calculated and compared with the amount of TCDD in the rat livers, it appeared to take less dioxin-like TEQs to induce the same enzyme activity in humans than in rats. In some other measures of dioxin-induced activity, the human tissue response was the same as that of the rat (Lucier, 1991).

There appears to be a real increase in cancers in older people which is not explained by an increase in the number of people surviving to older ages (Doyal et al., 1983). It is possible that some of this increase is connected with the rise in levels of dioxins and similar compounds in bodyfat over the last 40 to 50 years and their continued presence in the body for most of a lifetime.

The extreme persistence of dioxin in humans means that, once exposed, we continue to receive low doses from our bodyfat. Few animals have been studied for the equivalent of a human lifetime: can a rat's lifetime of two years be comparable to a human life expectancy of over 70 years?

Humans appear less sensitive than many animals studied because:

1 Less food is ingested per body mass 2 More TCDD is sequestered in adipose tissue and away from organs likely to be affected 3 Tissue sensitivity seems to be lower than in the most sensitive animals (Byard, 1987) 4 It was not known until fairly recently that everyone has some dioxin contamination, so there have been virtually no studies relating these low levels to ill health.

The ultimate effect of all these uncertainties may be that long term effects may not show up until it is too late.

Effects On The Foetus

The developing embryo or foetus in the womb does not share the adult human characteristics which may make effects of dioxins slow to become evident. Its susceptibility is more like that of the more sensitive animals for the following reasons:

1 It is growing more rapidly than at any time in later life. Cell division is occurring in a higher proportion of cells. The dioxin family affect this process very strongly. 2 The foetus lacks important drug-metabolising detoxification capacities that are found after birth (Jacobson et al., 1990). 3 Until near the time of birth the foetus lacks fat deposits that might dilute the impact of the exposure to pollutants (Jacobson et al., 1990). 4 The blood-brain barrier is incompletely developed so vulnerability to central nervous system damage is increased (Jacobson et al., 1990). 5 The small size means that intake of contaminants is disproportionately large. On the basis of milligrams per kilogram of bodyweight, the amount taken in by the embryo or foetus is greater than that of the breastfed baby (Rogan et al., 1988). 6 Brain development spans a long period pre-birth.

Pre-birth effects of the dioxin family include malformations, neurological effects and changes to the immune system perhaps giving rise to cancer or infections. In addition, there are effects via the father which are not related to actual intake by the baby, but either to changes in sperm or to events around conception, perhaps caused by chemicals attached to proteins carried in the sperm head (New York Times, 1991a).

Neurological and behavioural changes will be considered first, since these are possibly the most sensitive indications of the effects of the dioxin family, and they have been observed in babies born to mothers at the high end of "normal" levels in industrialised countries.

Neurological and behavioural disadvantage

There is ample evidence that PCBs, dioxins and furans cross the placenta in humans. This is confirmed by many different studies (eg. Van Wijnen, 1990 and Schecter, et al., 1990). It is now clear that passage of these compounds through the placenta has been the cause of damage that has occurred in human populations at levels which had little apparent effect on the adults.

As well as having many physical defects at birth, the children of women who had eaten PCB-contaminated cooking oil in Japan and Taiwan showed poorer performance on standardised intelligence tests when studied later (Rogan etal., 1988). These children in Japan and Taiwan were poisoned by contaminated cooking oil consumed by their mothers before they were born, and it is thought that the furans were responsible for the extreme toxicity, although toxic PCBs will have contributed (Tanabe, 1988). Affected children were still being born six years later. They were smaller than normal, they had discoloured skin and nails, abnormal teeth and gums, and many when studied later were apathetic and dull with IQs in the 70s. They also showed high rates of infections such as bronchitis (Rogan et al., 1988). The oil, because it had been heated, contained furans and other organochlorines as well as PCBs.

Researchers have looked for evidence of similar effects in normal populations which have low levels of PCBs and similar pollutants. A study of babies born to 313 mothers who had eaten PCB-contaminated fish from Lake Michigan in North America was started in 1980 (Jacobson et al., 1990). The mothers had eaten the equivalent of two or three lake trout or salmon meals per month over a period of about 16 years before the birth of the babies.

At birth, the most highly exposed babies were found to be 200-250 g lighter than relatively non-exposed control infants. This was valid when related to fish consumption by the mothers and when related to concentrations of PCBs in umbilical cord serum. The highly exposed babies had smaller head circumference and were born on average 6 to 12 days early (Swain, 1988).

Along with these more obvious physical differences, the babies also were examined at birth for more subtle behavioural deficits. 42% of exposed babies were classed as relatively unresponsive. 115 of the 313 had jerky uncoordinated movements in which flexor and extensor muscles appeared to be competing. Compared with unexposed babies, these newborns also showed greater tendency to startle, and a greater number of abnormally weak reflexes.

Further analysis showed that some babies, showing no physical effect such as low birth weight, still showed neurological differences. Some babies showed only physical effects.

Some of the babies were examined again at seven months. Their visual recognition memory was tested by measuring how long the babies looked at pairs of photos when a new photo was put beside a familiar one. This is a standard way of providing an index of attention to novelty, which is related to short-term visual memory. Infants who had the highest exposures before birth showed no preference for novelty, and the score on this test was clearly dependent on the PCB levels in umbilical cord serum.

Interestingly, the researchers found that the test score was related only to pre-birth exposure to PCBs; the amount of PCBs received via breastmilk made no difference.

A representative sample of 236 children from the Michigan study were followed up at age four (Jacobson et al., 1990). Cord serum PCB levels were available for 146 of the children; mothers who breastfed had provided samples of milk a few days after birth; serum samples were obtained for most of the children at age four, a few at age five. Samples were analysed for PBBs (Polybrominated Biphenyls - which contain bromine) as well as PCBs and the serum samples were also analysed for a range of organochlorine pesticides and solvents. DDT was the only pesticide found.

Children who had had higher intakes of PCBs before birth had lower scores on tests of various types of short-term memory than controls. There was no gross impairment, but definite evidence of diminished potential.

Since PCBs and similar contaminants have been found in breastmilk, fears have been expressed that the intake through breastfeeding could cause ill effects. However, in this study, the effects depended only on the intake of PCBs before birth, even though the intake through breastfeeding is much larger. Breastfeeding was shown to cause high levels in the serum at age four, but the memory test results were found to be independent of serum levels. In fact, in all but the lowest PCB levels of cord serum studied, the scores on the tests rose with the duration of breastfeeding. Breastfeeding had a positive effect in spite of its pollutant load and this effect was lessened but still remained after accounting for various socio-economic factors such as quality of home, maternal education, etc.

The fact that effects caused pre-birth were still seen in these children up to the age of four is alarming. It is possible that such effects could persist for life, and this is suggested by the evidence from monkeys exposed to PCBs through their mothers (Schantz, 1991). Facilitation of learning was observed at 4 to 6 years (young adult monkeys) from low exposure, and higher exposure caused a decrease in learning ability. The authors say this is a similar pattern to that caused by lead - a decrease in attentiveness makes it possible to ignore weak stimuli and hence concentrate on the main task better, but a stronger decrease causes learning problems. A suggestion of similar long term changes is also seen in the TCDD exposed monkeys studied earlier (Bowman, 1990).

The authors of the Michigan study ascribe the effect of breastfeeding to the greater stimulation it provides for the child; it may also be due to other factors such as nutrients in the breastmilk (Bitman, 1986). In addition, 17 children who had similar serum levels to the others, both at birth and at four years, but whose mother's breastmilk was more contaminated, refused altogether to cooperate with the development tests; the research does not make it clear why this was, or how much they had been breastfed (Jacobson et al., 1990).

Research carried out as part of the U.S. North Carolina Breastmilk and Formula Project confirms the Michigan results. Nine hundred and thirty babies born between 1978 and 1982 were studied from birth until 1986 (Rogan et al., 1986).

The study examined the relationship of exposure to PCBs and DDE before birth to various neonatal effects. It found that birth weight, head circumference and neonatal jaundice were not related to PCBs or DDE, but that poor muscle tone and poor reflexes were associated with higher PCB levels. Higher DDE levels showed a correlation with poor reflexes. Again the effects were independent of the amount of breastfeeding: they depended only on the amount received before birth.

The babies were also examined at six months and twelve months, using standard tests of infant development (Bailey scores). Higher exposure to PCBs before birth was associated with lower scores on the psychomotor scale. A similar measure of mental development showed no effect of PCBs.

These effects occurred at the "high end of normal background PCB", levels seen in around 5% of the sample (Tilson et al., 1990). A small change in psychomotor development was seen with each increment in PCB levels. At six months this index dropped 0.96 point for every increase of 1 ppm in PCBs in milk fat, which was calculated as an index; at 12 months the drop was estimated at 1.34 points per ppm.

PCBs consist of a mixture of different congeners and isomers. 209 are possible altogether. It is difficult to compare PCB levels in different countries because different mixtures have been used in various regions and therefore the method of analysis involves comparison with an industrial standard. Michigan levels were analysed in the early 80s using a type of analysis similar to that employed in the UK. Means of 0.74 ppm and 0.64 ppm were found, compared with a mean of 0.5 ppm found for UK breastmilk in 1982/83 (Jensen, 1987). The North Carolina researchers used a method which tends to give a higher result than the two similar methods above. Hence UK levels should be compared with the Michigan levels rather than with North Carolina and conclusions about threshold levels for effects drawn from the Michigan research can reasonably be applied to this country. The UK government state that levels in the UK appear to be lower than in many other countries (MAFF, 1989) but they are not, in fact, much lower than the levels in Michigan.

However, all the PCB analyses quoted here for the two US studies and the UK survey are much less accurate than the best modern method of analysing separately for each important congener. Since it is now known that the mono ortho coplanar PCBs are the main contributors to toxicity, what really should be reported is the TEQs from these, plus those from dioxins and furans. It would then be possible to relate the pre-birth effects to actual levels. The use of the old methods for the North Carolina and Michigan studies is a weakness, but since both studies agree in many respects, and are otherwise well researched, it would be wrong to ignore them. It may well turn out that the total PCB levels do provide a good guide to the effects of the total dioxin-like toxicity.

Both the Michigan and North Carolina studies assumed initially that there was likely to be no threshold level for effects (Tilson et al., 1990). That means it was assumed that even the lowest levels were likely to have some effects. However because the effects of dioxin-like compounds depend on the Ah receptor, and there has to be a certain (very low) concentration in the cells before binding to the receptor will occur, there could be expected to be some low level which would cause no effects. Accordingly, researchers from Michigan and North Carolina got together (Tilson et al., 1990) to look in the results for levels at which effects appeared to start. They also agreed on an approximate "translation" between the PCB levels measured with the two different methods of analysis used in North Carolina and Michigan, so that levels can be compared.

Some 'threshold' levels for pre-birth neurological effects of PCBs on babies have been estimated (Tilson et al., 1990). For the batteries of standard tests used, the level appears to be about 3.4 ppm of PCBs in fat, but it is as low as 1.8 ppm for the method of PCB analysis used in the Michigan Study. For visual recognition memory alone, the level is 1.0 ppm in fat (measured as breastmilk fat). They go on to calculate "acceptable daily intakes" but it is possibly less confusing to look at actual bodyfat levels, since these dictate the intake to the baby in the womb. Some effect on visual recognition memory may be occurring at as low as 1 ppm in breastmilk fat, according to these researchers' examination of the Michigan study. This is a level at which effects are found. The authors suggest a safety factor of 10, and note that this is quite small. They point out that the number of Ah receptors can vary by 7 fold and that this factor is not alone in determining susceptibility to dioxin-like compounds. This argument omits other effects of PCBs not acting through the Ah receptor which may affect brain function (Seegal et al., 1990), so a factor of 10 may be too small. The researchers imply that a "safe" level of PCBs in maternal bodyfat is 0.1 ppm. However, Tilson's paper tends to assume that all the effects in the Michigan and North Carolina studies were due to PCBs: it is likely that dioxins, furans and other similar compounds contributed as well.

Levels of PCBs in bodyfat were measured in the UK in 1976 and 1982 (MAFF, 1982; 1986). The proportion of samples containing over 1 mg/kg (ppm) increased from 16% in 1976 to 34% in 1982. Out of a sample of 105 males, the range was 0.1 - 6.9 ppm, and the mean was 1.0 ppm. Out of 82 women the range was from undetectable to 2.2 ppm, with a mean of 0.8 ppm.

Bodyfat has not been measured recently, although breastmilk has, and the results will probably be published in 1992. Levels in breastmilk fat are closely related to the amount in bodyfat, and so give a guide to the likely intake of the baby before birth. Levels in breastmilk fat in 1980 were between under 0.1 and 2.1 mg/kg (Collins et al., 1982). The mean is 0.5 ppm; nearly half of the mothers were between 0.5 and 1.0 ppm. About 8% were over 1.0 ppm. If we are similar in our body levels of the dioxin-like compounds to North Carolina, then effects could be seen at 1.8 ppm and above. This was 5% of the North Carolina sample. 3.3% in North Carolina had levels over the UK maximum of 2.1 ppm, so, on this basis, the percentage at risk in the UK could be 1-2%.

Levels of dioxins and furans in breastmilk in the UK are 29 and 37 ppt TCDD toxicity equivalents (TEQs) in Glasgow and Birmingham (Startin et al.,1990). These are higher than most US levels: in Tennessee, an adjacent Southern State, similar to North Carolina, the level is 14.6 ppt TEQs (Schecter et al., 1990b).

The above arguments about threshold levels suggest that effects are likely to be found, but they should not be taken at face value. In spite of improvements in analytical techniques, results of analysis for dioxins, furans and PCBs can differ by factors as great as 2 between different laboratories. It is likely that the 'translation' between the Michigan and North Carolina results (Tilson et al., 1990) is validated between the two laboratories involved, but other comparisons have to be treated with caution. This means that the levels we are discussing where effects could be found are within the margin of error between laboratories. Clearly this means that in the UK we could either be very clearly within the range at which effects could be seen, or we could be out of it. Since we do not know, we must behave as if children are likely to be affected and act to remove the risk by reducing our levels a great deal. The laboratory error margin makes the risk we are taking with nearness to thresholds unacceptable.

Dioxins have also been shown to cause pre-birth effects in experiments on monkeys.

2,3,7,8-TCDD, the most toxic dioxin, causes changes in mother/infant interaction and possibly decreased visual attention in rhesus monkey babies, at the lowest intakes measured, ie. 120 pg per kg bodyweight per day before conception (Bowman, 1989a). Differences in maternal care were seen when monkey mothers had this intake daily for 45-49 months and then a dioxin-free diet for 10-12 months before mating.

The monkeys born after this period were given a battery of tests, some of which were similar to those carried out on the Michigan babies. The most striking effect was on the interaction between mother and baby. The TCDD infants spent more time closely cuddling their mothers and suckling. They left their mothers less, and when they did leave, their mothers fetched them back more quickly than normal. This was analysed statistically: the mother-baby distance was less than in controls. Similar effects had been observed where baby monkeys had high lead intakes, and these are known to be effects via the baby rather than the mother, because of fostering studies. It is as if something about the baby makes the mother keep it closer to her. Similar behaviour is observed when monkey babies are ill or injured; these babies, however, showed no sign of illness. The observed effects might have been even more noticeable if more of the baby monkeys had been female: the exposed group consisted of 2 males and 5 females, and it was compared with a control group of 4 females and 2 males. Monkey mothers tend to allow a greater distance from male babies.

Neonatal effects were also observed: the babies were "more passive at neonatal assessment". This is evidence that a pre-birth effect is involved. The effects were rather like those found in North Carolina in babies exposed to PCBs. The monkeys were also found, between 2 and 4 months old, to have a decreased visual attention - they showed less curiosity in looking at a series of slides than did the controls. This is reminiscent of the effects of PCBs in the Michigan babies, some of whom showed no attention to novelty at seven months.

The researchers have been trying to relate these effects to the bodyfat level in the monkey babies at weaning, in spite of the evidence for a pre-birth effect, as seen in the neonatal effects. Thus they have not investigated how the effects vary with the level in the mother's fat at birth. However they do give levels for TCDD in the mother's fat for the low exposure group and for the higher exposure mothers. The lower levels are an average of 49 + 11 ppt TCDD in the mother's fat. This is only slightly above the average TEQ levels found in the breast milk fat of mothers in Birmingham and Glasgow, and it is a level at which noticeable effects occur in monkey babies. This level is erroneously described as a no-effect level in the UK government's report (DOE, 1989) but the same report also says "there would seem to be no good reason for setting aside the monkey data in any risk assessment of TCDD for humans".

There is a difference in the relative concentrations of dioxins in mother and baby between monkeys and humans and this could be used to suggest that the results of the monkey experiments are not applicable to humans. The average concentrations of TCDD in the monkey babies' fat is approximately three times that of their mothers, whereas in human babies at birth it is much lower than in adults (Ryan, 1986; Beck, 1990). However, human babies may be much more similar to monkey babies when in the womb before the fat has developed. The differences at birth may not be at all relevant to pre-birth effects. Thus some effects are seen in monkeys at levels of 2,3,7,8-TCDD which must be fairly common in human mothers, if the contribution from different dioxins and furans to total TEQs is considered.

The monkey babies may not have received any more dioxins from their mothers than human babies would, but because they have less fat at birth, they have higher concentrations in their fat. The intake to vital organs and the nervous system before birth could be similar in both species at the same levels in bodyfat.

Thus, from dioxins and furans alone, average levels in bodyfat in this country are close to the levels which cause effects in monkey babies. Many individuals in the UK are likely to be at or above these levels.

In both US studies we are probably seeing the combined effects of PCBs with other members of the dioxin family. The Michigan researchers controlled for effects of other contaminants such as lead and pesticides. They could not analyse such a large sample for dioxins and furans meaningfully, so could not distinguish between the contributions of these different pollutants. Consequently, PCBs were acting as a marker for total dioxin-like toxicity.

The PCB mixtures may be different from study to study around the world, with varying amounts of the toxic congeners producing different toxicities in different areas. This causes uncertainties when making comparisons between different studies.

Furthermore, different methods of measurement and analysis may be so variable between studies that comparisons cannot be made. However, an approximate translation can be made between the US and UK results, as the UK study used a similar method of measurement to the Michigan study.

It is important to stress that these uncertainties could just as easily work in our favour as against us.

If total PCBs are used as marker for the other compounds, and assuming our total levels of dioxin-like toxicity bear a similar relationship to the levels of PCBs to that in Michigan and North Carolina, we can assume that some neurological effects will be seen here.

If we are similar to North Carolina in our levels of dioxin-like toxicity, with effects beginning at 1.8 ppm PCBs, then there could be effects in 1-2% of the UK child population.

If our levels of dioxin-like compounds are more like those in Michigan, then effects could start at 1.0 ppm of PCBs. This would mean that a much larger percentage would be affected, perhaps 8 %, since about 8 % of mothers tested in 1980 had levels over 1.0 ppm.

In fact, our levels of dioxins and furans are probably somewhere in between the Michigan and North Carolina levels and in some areas they may be as high as those of fish eaters in Michigan.

In addition, whilst evidence from experiments on monkeys may not compare exactly with effects in humans, it is likely to be of a similar order of magnitude. Maternal bodyfat levels of total dioxin and furan toxicity are similar to those known to cause behavioural changes in monkey babies. At the 1991 Dioxin Symposium, held in North Carolina, it was estimated that the toxic non-ortho and mono-ortho coplanar PCBs could double or triple the toxicity from dioxins and furans. Applied to the UK, this means that our levels of toxicity from dioxins, furans and co-planar PCBs is perhaps double the amount from dioxins and furans alone, at about 70 ppt TEQs. This is higher than the levels which cause effects in baby monkeys.

There may also be additional small contributions from chlorinated naphthalenes.

It is likely therefore, that a small percentage of children in the UK will be affected by levels of dioxin-like compounds.

People who live in an area where dioxin intake is relatively high and where fish consumption is high are more likely to see effects in their children.

Even if dioxin-like compounds affect only 1 % of the population, this means thousands of children with what the Michigan researchers referred to as "diminished potential".

Effects Through the Father

The dioxin family has effects on reproduction which depend on the father's contact with these substances. It has been suggested that the reduced sperm count in industrialised countries may be connected with organochlorines, and there is considerable evidence of other effects connected with sperm (Dougherty et al, 1980).

When couples are infertile, this is due to the man in about half the cases. It is regarded as a random misfortune, but there is evidence that organochlorines may be to blame, as well as other toxic substances. Environmental chemicals may be to blame in half the cases of male infertility where there is no obvious cause (Dixon, 1979, quoted in Pines,1987).

Some chemicals, like radioactivity, cause mutations; that is, they alter genes. It is thought that the dioxin family are cancer promoters, rather than mutagens: however, they are always present in the body with other carcinogens and are known to make other substances carcinogenic (Silbergeld & Gasiewicz, 1989).

Vietnam veterans who were exposed to 2,3,7,8-TCDD (in Agent Orange) have a lower sperm count than controls (AOTF, 1990). The Agent Orange Task Force (AOTF), carrying out an independent study, concluded from all the evidence that reproductive and developmental effects in the families of veterans were "at least as likely as not" to have been caused by exposure to the dioxins in Agent Orange. These effects include low sperm count, and increased incidence of miscarriage for the wives of the veterans.

Research in Israel has suggested that infertile men have higher levels of PCBs and organochlorine pesticides in their blood than fertile controls Pines, 1987). However the researchers found that organochlorine levels now (1987) were lower than ten years before. Most of the men examined were born in the 1950s and 1960s, when pesticide and PCB use were at their height. They would have had high intakes before birth and as babies. This has been shown to cause sterility in animal experiments.

There is evidence that in the US and other industrialised countries, sperm density has fallen since 1950 (Dougherty et al, 1980). The range found between 1929 and 1951 was between 100 and 145 million sperm per cubic centimetre. The average found in studies after 1970 ranges between 48 and 80 million per cubic centimetre. Sperm density decreases with the number of cigarettes smoked per day, and the number of abnormal sperm cells increases. Other mutagens have similar effects. Sperm are constantly produced by cell division, and substances which inhibit cell division will reduce sperm count. Most of the known carcinogenic substances have this effect (Dougherty et al, 1980).

The fall in sperm density is particularly dramatic in the late 70s and in 1980. A man with less than 20 million sperm per millilitre is regarded as functionally sterile. The number of men in this category has risen since sperm density started to be measured, and it has risen especially steeply since 1950. Research in 1975 and 1980 showed that low sperm levels were strongly associated with toxic organic contaminants in the seminal fluid. However, only very low levels of dioxins and furans are likely to be present in semen. Research in Canada failed to detect any (Ryan,1986). Thus any effects, if caused by dioxins, are likely to be dependent on body levels of dioxins rather than on levels in the testes. For other chemicals, present in greater concentrations in all organs, the levels in semen may be an indicator, rather than a direct cause.

In 1938, only 0.5% of men had a sperm count lower than 20 million; now the percentage is much higher, possibly 20% (Elkington, 1985)

Thus there is the possibility that the inability to father children is due both to present exposure and to damage caused around the time of birth.

Miscarriages and congenital malformation via the father

There has been considerable dispute about the effects of Agent Orange. Many official studies have found no evidence for the various health effects, but a recent independent study by eminent scientists considered all the evidence, and concluded that there was a significant statistical association between several serious illnesses and exposure to Agent Orange. Reproductive effects are among other effects which "are at least as likely as not" to have been caused by Agent Orange. This statement is significant because it is the criterion used by the US Department of Veterans' Affairs to decide if compensation can be paid in cases of ill effects due to chemical exposure.

The AOTF also list all the research papers they studied and publish lists of those studied by the government and those studied by the Task Force, but not by the government.

Two US studies are quoted that found significantly higher rates of miscarriage in spouses of Vietnam veterans exposed to Agent Orange (AOTF, 1990). Greater foetal loss is also reported in spouses of veterans in Tasmania. Studies in North Vietnam suggested that male exposure to TCDD in South Vietnam during the war caused higher rates of miscarriage and congenital malformation several years later (AOTF, 1990).

The major birth defects found in children of Vietnam veterans include malformations of a wide variety of organs including the nervous system (eg. hydrocephalus), the heart, genital organs, and urinary tract. Other major effects include cleft palate, club foot, and hand and limb deformities (Silbergeld & Gasiewicz, 1989; Birnbaum et al., 1985). Some studies found increased risks of spina bifida and congenital cancer. An Australian study found some evidence of increased heart defects and Down's Syndrome. Other studies showed evidence of learning difficulties perhaps reflecting subtle nervous system defects (AOTF, 1990). More minor defects included abnormal palm creases, and dimples in skin at the base of the spine (Aschengrau & Monson, 1990). Some studies have found excesses of birth marks and skin discolorations.

The children of Vietnam veterans have increased incidence of skin defects, nerve defects, heart defects, kidney defects and cleft lip and cleft palate. The Agent Orange Task Force considers that these effects are "at least as likely as not" to have been caused by exposure to Agent Orange.

Studies in North Vietnam, of fathers who had served in South Vietnam during the war, found that the rates of miscarriage and congenital malformation were raised, and that there were more of these cases the higher the father's exposure, as indicated by the part of South Vietnam in which he had served. These effects persisted for many years after exposure. Some other reproductive effects did not appear to be related to the father's exposure. Possibly these depend mainly on the mother's contact with Agent Orange.

It is not possible to say that organochlorines are the only cause of lowered sperm count - all kinds of different factors can do this. However, the Agent Orange and PCB evidence suggests that the dioxin family could cause sterility, and the rise in levels of these compounds since World War II parallels the decrease in sperm count observed in the USA.

Several childhood cancers have been linked to new mutations in sperm, not eggs (New York Times, 1991b). Cells are more vulnerable to damage when they are dividing in order to produce more. When a cell divides, a new copy of DNA is made; this copy is likely to be imperfect if a mutagen interferes. Cells which are dividing are more likely to absorb toxic substances than cells that are not. New sperm are produced constantly by cell division; egg cells are formed in the womb before birth. Thus damage to sperm happens more easily, although its effects may disappear when the toxic substance is removed. The dioxin family remains in bodyfat for many years, and dioxins make other chemicals mutagenic. Thus a man with a high level of dioxin in his bodyfat may have a long-term risk of fathering children with cancer because of an enhanced mutation risk from other substances he comes into contact with.

Recently, it has been shown (Cartwright, 1991) that a father's occupation is associated with childhood leukaemia. This was statistically significant for father's exposure to wood dust before the child's conception; also for exposure to benzene, as well as to radiation, in this period. Working in the chemical industry is also associated with a doubled risk of fathering a child with leukaemia (Gardner, 1990).

Wood dust is recognised as a carcinogen: this effect is likely to be enhanced (Hardell, 1983) because it is treated with wood preservative, often pentachlorophenol which contains dioxins. There could be a combined effect of reducing sperm count and causing defective sperm. Benzene is a known carcinogen and could have similar effects.

Effects of dioxin have occurred via the father in Vietnam veterans, although the mechanism is not known. There is some possibility that dioxins and PCBs affect male fertility and that chemicals can directly alter sperm. There is evidence from this country that the occupation of the father is linked with serious illness in children, although again, the mechanism is not known. However, it appears likely that the father's exposure to dioxin-like compounds could be adding to health risks associated with levels of these compounds in the mother's body.

Women

Women suffer from the effects of the dioxin family on men (AOTF, 1990). For example, wives of Vietnam veterans suffered more miscarriages. However, there are effects which clearly arise because of direct effects on the female reproductive system.

Health in Pregnancy

Mothers who ate PCB contaminated fish from Lake Michigan suffered direct effects on their pregnancies (Swain, 1988). Higher fish consumption was associated with anaemia before and during pregnancy, water retention and swelling, and an increased rate of all kinds of infections. Toxaemia of pregnancy, an occasionally fatal condition, has been associated with high blood serum levels of PCBs (Wassermann et al., 1980)

It is thought (Wasserman et al., 1980) that pregnancy toxaemia is due to the mother's poor immune response to cells originating from the baby. Antibodies do not deal with them, and their continued presence then provokes a massive reaction, with raised blood pressure and fits. Since high PCB levels are found in some cases of toxaemia, they may cause it by affecting the immune system. The immune system can be affected by many different factors such as diet. However, there is other evidence for effects of PCBs on the immune system, in the persistent infections of Yusho patients, and in the pregnancy illnesses of women in the Michigan study (Swain, 1988).

Miscarriages have been linked with higher levels of PCBs: in Italy a small study found that women who had miscarriages had more PCBs in their bodies; especially congeners with 4 and 5 chlorine atoms (Leoni, 1989)

Women exposed to TCDD in the Seveso accident in 1976 had abortions because of fears of ill effects. The foetuses had higher levels of chromosome aberrations (Tenchini et al., 1983) than those from non-exposed women who had abortions for other reasons.

The neurological and behavioural effects of PCBs in North Carolina and Michigan are clearly related to the mother's intake since the effects are greater at higher levels of PCBs in the mother's bodyfat.

In Vietnam, because the whole environment was contaminated with Agent Orange, husbands and wives were both exposed to TCDD. Several effects on pregnancy and birth, including hydatidiform mole, are thought to be linked to dioxin exposure. Hydatidiform mole is an abnormal pregnancy in which placental tissue grows in an uncontrolled way and the embryo does not develop. It carries a risk of cancer which can be fatal without intensive chemotherapy. There is a suggestion that levels rose steeply in 1976, compared with incidence in the 1950s and 1960s (Nguyen Thi Ngoc Phuong (1), 1989). It is more common in South-East Asia than in other parts of the world, so it is possible that incidence in Vietnam was simply noticed more after the spraying by Agent Orange, however, some of the cases seen at a hospital in Ho Chi Minh City in 1982, do seem to be associated with spraying.

A large number of horrifically deformed babies have been born in Vietnam since Agent Orange was used there. It cannot be doubted that these are associated with the spraying, but the isolation of Vietnam has meant that there are very few studies which strongly support this. The above study also links the incidence of deformity with intensity of spraying.

Infant mortality has also been shown to be raised in South Vietnam (Le cao Dai, 1990). A proportion of deaths of children before the age of one year is due in all countries to congenital malformations, and to other factors apparent at birth, such as prematurity. It was chosen by the researchers because it was a simple occurrence to survey, easier perhaps than stillbirth or deformity (it was estimated in Glasgow that 50% of infant deaths after the neonatal period were due to factors apparent at birth - (Macfarlane & Mugford, 1984)). Hence, it is likely that it reflects levels of malformation. There were more deaths in areas that were heavily sprayed. The relative risk was 2.4 in 1966-1970 for the heavily exposed villages relative to a very similar unexposed control village. The rate of infant mortality remained fairly consistent in the control village over the whole period to 1980. There was a statistically significant excess of infant mortality in the sprayed villages until 1980, and a slight excess between 1981 and 1986. The rate of infant mortality was 6% in 1966-70. There were very high levels of 2,3,7,8- TCDD in breastmilk when spraying stopped in 1971. Samples taken between 1970 and 1973 contained up to 1450 pg of TCDD per gram of fat. Levels in the two sprayed villages were still raised in 1985 to 1987.

2,3,7,8-TCDD and TCDD Equivalents in Human Breast Milk from Three Villages in South Vietnam (from Le cao Dai, 1990).

Village 2,3,7,8-TCDD (pg/g lipid) Nordic TCDD Equivalent pg/g)

Can Gio 9.0 13.2

Tan Uyen 6.4 19.5

My Thanh 2.0 6.8

The third village was unsprayed. Its levels are much lower than in the other villages, and also much lower than levels found in industrialised countries. The other village levels are nearer to industrial levels: the TCDD levels are comparable but high, and the total TEQs lower.

The researchers suggest that the fall in the rates of infant mortality since 1966-70 parallels the fall in the TCDD levels and that this "supports the idea that TCDD may be a causative agent". There is still a slightly raised risk (1.3) in 1981-1986, although this is not statistically significant. It represent a 30% extra risk of a child dying before the age of one year for villagers in the villages which were sprayed before 1971.

North and South Vietnam, which differ markedly in exposure to dioxins, offer an opportunity to compare effects on babies and children which has not been fully utilised (Schecter et al., 1986). There have been some studies, but nothing which does justice to the suffering in Vietnam or to the importance of this question to people all over the world.

In addition to the effects in Vietnam, there is a litany of cases of birth defects throughout the world which are thought to be linked with exposure to the dioxin family of compounds. For example, babies born to mothers who suffered in the two poisoning incidents in Japan and Taiwan in 1968 and 1979 had severe physical defects, including cola-coloured skin, discoloured nails and abnormal teeth and gums, as well as the mental effects mentioned earlier. The connections of such cases with dioxin exposure have to be regarded as tentative however since we have only limited details of the amounts of dioxin thought to be responsible. Also, the number of such cases is thankfully small, so it is difficult to analyse them statistically.

In this country there are suspicions regarding the number of eye defect complaints amongst inhabitants in the areas around incinerators burning PCBs. There is the possibility of exposure to unburnt PCBs in such areas, as well as to dioxins, furans and other products of incomplete combustion. The eye defects, which are normally extremely rare, include a considerable reduction in size and absence of eyes at birth. According to press reports, four children were born in the Pontypool area of South Wales between 1980 and 1984 with such a class of eye defect. Five babies with similarly rare eye defects were reported to have been born within 13 months of each other at Bonnybridge, near Stirling in Scotland. Also, eye defects were reported around the Beridge incinerator in Nottingham. This has since closed down but the equipment is still in use near Sheffield (Greenpeace, 1991). However, official studies have failed to find eye defects in these areas (Gatrell & Lovett, 1990) and this may be because OPCS (Office of Population Censuses and Surveys) figures rely on voluntary notifications made by District Health Officers.

There is now a prospective study of eye malformations in progress at Moorfields Eye Hospital in London (Gatrell & Lovett, 1990). This means that some of the problems of under-reporting which have occurred in the past will be mitigated.

At least one PCB incinerator has been associated with raised levels of dioxins, furans and PCBs (Welsh Office, 1991). There is evidence of synergistic effects between 2,3,7,8-TCDD and one PCB congener in causing birth defects (see earlier).

Other Major Birth Defects

Major defects of the central nervous system such as spina bifida and hydrocephalus have been found in animals dosed with TCDD. There are many cases where human exposure to dioxins, especially from 245-T, has been linked with this type of birth defect, but the individual studies have tended to be dismissed as statistically insignificant. For instance, TCDD-contaminated oil was sprayed onto roads in Missouri, US, and mothers who were thought to be exposed during pregnancy were studied. The results appear to show that there was no statistically significant increased risk of infant death or birth defect, although the relative risks of defects of the central nervous system do appear to be raised and showed the greatest risk ratio (3.0). The study was limited because there was no actual measurement of body levels for any of the women. Some of them may not have been exposed at all (Hoffmann et al., 1989).

Intakes Before Birth in the U.K.

For the intake to the foetus, we need to consider levels actually in the mother's bodyfat. Levels in blood serum fat are similar to those in fatty tissue; it is the blood serum fat level and the extent to which each compound passes through the placenta which governs the intake to the foetus.

There is evidence that some dioxins and furans cross the placenta more readily than others. In marmoset monkeys 2,3,7,8-TCDD will do so when the concentrations on each side are in a ratio of 2 to 1; 23478-PeCDF requires a relative concentration of 10 to 1, and therefore crosses less readily (Krowke et al.,1990). Thus it is possible that the level of 2,3,7,8-TCDD is more significant for the foetus than the total TEQs. However, again, very little is known about how readily all the different dioxin-like compounds, such as coplanar PCBs, cross the placenta, except that there is strong evidence that most do to some extent.

The dioxin intake of a breastfed baby is approximately 120 pg/kg bw/day TEQs from dioxins and furans in the UK.

The levels which caused liver changes in rats in the Kociba study (Kociba, 1978) were 1000 pg/kg bw/day. So the intakes of breastfed babies are of the order of a tenth of those of the rats in the Kociba study.

One reason why the babies are not poisoned is that the intake goes into fat rather than to target organs like the liver. Fat is being increased all the time, which will increase the effect of this (WHO, 1990). However, if the intake before birth per kilogramme of bodyweight is higher than 120 pg, then unborn babies are receiving comparable levels to the Kociba rats, without the possibility of fat storage that the baby has after birth. It is far more likely that effects will be caused by the intake before birth than from breastfeeding.

8/10/96


Gascape's Reference Database© | Index | Home

©