To:
Subject: / 05ART/9094
Message Date: 5/2/2006 08:02
From: Bryan, Rose (ELS) <>
To: '' <
Dear Professor Grandjean,
I am delighted to be able to give you the news that your paper on
Developmental toxicity of industrial chemicals has now been accepted for
publication in The Lancet as a Review.
The next step is that you will receive proofs from one of my Assistant
Editor colleagues. Please note that we do substantially subedit manuscripts
to improve clarity and to our house style.
Many thanks for all your hard work on this Review.
Regards.
Rose Bryan
sent on behalf of Dr Astrid James
Deputy Editor - The Lancet
Tel: 020 7424 4921
Fax: 020 7424 4912
Manuscript submitted to The Lancet (accepted for publication, 2 May 2006)
Developmental neurotoxicity of industrial chemicals –
A silent pandemic
P Grandjean, PJ Landrigan
Institute of Public Health, University of Southern Denmark, 5000 Odense, Denmark, and Department of Environmental Health, Harvard School of Public Health, Boston, MA, USA (Prof P Grandjean MD DMSc)
Departments of Community Medicine and Pediatrics, Mount Sinai School of Medicine, New York, NY, USA
(Prof PJ Landrigan MD MPH)
Correspondence to:
Dr Philippe Grandjean, Department of Environmental Health, Harvard School of Public Health, Landmark Center 3E-110, 401 Park Drive, Boston, MA 02215, USA. Telephone: +1-617-384-8907. Fax: +1-617-384-8994. E-mail:
Abstract
Neurodevelopmental disorders (NDDs), such as mental retardation, attention deficit disorder, cerebral palsy, and autismare common, costly and can cause lifelong disability. Their aetiology is mostly unknown. A small number of industrial chemicals – lead, methylmercury, polychlorinated biphenyls, arsenic, and toluene – are recognised causes of NDDs and of subclinical brain dysfunction. Exposures to these chemicals during early development can cause brain injury at dose levels much lower than those affecting adult brain function. Recognition of these risks led to evidence-based programs of prevention, such as eliminating lead additives in petrol. Although already highly successful, these prevention campaigns were initiated only after substantial delays. Two hundred additional chemicals are known to cause clinical neurotoxicity in adults. Despite a lack of systematic testing, many more are known to be neurotoxic in laboratory models. Their toxicity to the developing human brain is not known and they are not regulated to protect children. The two major impediments to preventing neurodevelopmental deficits of chemical origin are the great gaps in testing chemicals for developmental neurotoxicity; and the high level of proof required for regulation. New, precautionary approaches are required for chemical testing and control that recognise the unique vulnerability of the developing brain.
Search strategy
We identified industrial chemicals that have caused neurotoxic effects in humans from the Hazardous Substances Data Bank of the National Library of Medicine, as supplemented by fact sheets at the (US) Agency for Toxic Substances and Disease Registry, and the Integrated Risk Information System of the US Environmental Protection Agency. We used “neurotoxic”, “neurologic”, and “neuro-” as search terms. For each neurotoxic substance identified, we then used synonyms, commercial names and CAS numbers to search PubMed, TOXNET, and TOXLINE to identify literature on developmental neurotoxicity. The primary search terms were “Prenatal Exposure Delayed Effects”[MeSH] and “Neurotoxicity Syndromes”[MeSH]. Secondary searches used combinations of “maternal exposure” and “maternal fetal exchange” with “developmental disabilities/chemically induced” and “neurotoxins”, all of them with the limiters “All Child: 0-18 years, most recent 10 Years, English, Human”. We also used references cited in the publications retrieved.
One out of every six children has a developmental disability, and in most cases these disabilities involve the nervous system.1 The most common neurodevelopmental disorders (NDDs) include learning disabilities, sensory deficits, developmental delays, and cerebral palsy.1 Some observers report that prevalence of certain NDDs – attention deficit/hyperactivity disorder and autism, in particular – may be increasing, but data to sustain that position are limited.2 The treatment of these disorders is difficult, and the disabilities they cause can be permanent.3 They are very costly to families and to society.4-6
Evidence has been accumulating over several decades that industrial chemicals can cause NDDs, and that subclinical stages of these conditions may be common. The suspicion of a link between chemicals and widespread neurobehavioural damage was first raised by research demonstrating that lead is toxic to the developing brain across a wide range of exposures.7-10 It was buttressed by reports indicating that other environmental pollutants are also toxic to early brain development.11 A report from the US National Research Council concluded that 3% of developmental disabilities are the direct consequence of neurotoxic environmental exposures, and that another 25% arise through interplay between environmental factors (chemical and other) and individual genetic susceptibility.3These estimates were based on the limited neurotoxicity information available and may therefore underestimate the true prevalence of chemically-induced abnormalities.
Neurobehavioural damage caused by industrial chemicals is, in theory, preventable. An essential prerequisite to prevention is recognition of a chemical’s ability to harm the developing brain. Knowing that a chemical is neurotoxic can trigger efforts to restrict its use and control exposures.Previous evidence-based programs of exposure prevention, such as those directed against children’s exposure to lead, have ultimately been highly successful, although they were initiated only after substantial delay.
The purpose of this review is to characterise the vulnerability of the developing nervous system to chemical toxicity; to tabulate publicly available data on human neurotoxicity of industrial chemicals; to examine the possible extent of a developmental neurotoxicity pandemic; to describe the known consequences of developmental neurotoxicity at individual and societal levels; to examine the implications for human health of the current lack of toxicological information; and to consider prospects for prevention.
Vulnerability of the developing brain
The developing human brain is inherently much more susceptible to injury caused by toxic agents than the brain of an adult.12 This susceptibility reflects the fact that in the nine months of prenatal life the human brain must evolve from a strip of cells along the dorsal ectoderm into a complex organ comprised of billions of precisely located, highly interconnected and specialised cells. Optimal brain development requires that neurons move along precise pathways from their points of origin to their assigned locations, that they establish connections with other cells near and distant, and that they learn to intercommunicate.12-14All of these processes must take place within a tightly controlled time frame, in which each developmental stage must be reached on schedule and in the correct sequence. Because of the extraordinary complexity of human brain development, windows of unique susceptibility to toxic interference occur that have no counterpart in the mature brain, or in any other organ. If a developmental process in the brain is halted or inhibited, there is little potential for later repair, and the consequences may then be permanent.12,14
During foetal development, the placenta offers some protection against unwanted chemical exposures, but it is not an effective barrier against environmental neurotoxicants.15 For example, many metals easily cross the placenta, and the mercury concentration in cord blood is considerably higher than in maternal blood.16 The blood-brain barrier, which protects the adult brain from many toxic agents, is not completely formed until about 6 months after birth.17
Postnatally, the human brain continues to develop, and the period of heightened vulnerability therefore extends over many months through infancy and into early childhood. While most neurons have been formed by the time of birth, growth of glial cells and myelinisation of axons continue for several years.13,14
The susceptibility of infants and children to industrial chemicals is further amplified by their increased exposures, augmented absorption rates, and diminished ability to detoxify many exogenous compounds relative to adults 18, 19 Persistent lipophilic substances, including certain pesticides and halogenated industrial compounds, such as polychlorinated biphenyls, accumulate in maternal adipose tissue and are passed on to the infant via human milk, thereby resulting in exposures to the infant that exceed the mother’s own dietary exposure by 100-fold on a body weight basis.20
Recognition of developmental and subclinical neurotoxicity of chemicals
Initial clinical recognition of developmental neurotoxicity in children exposed to industrial chemicals has typically involved obvious functional abnormalities in high-dose scenarios that caused obvious poisoning. Subsequent, more sophisticated research has documented the presence of less striking, but nonetheless serious adverse effects at lower levels of exposure (figure 1). This sequence of discovery led to the recognition that environmental toxicants exert a range of adverse effects – some are clinically evident, but others can be discerned only through special testing and are not evident on the standard examination, hence the term “subclinical toxicity.” The underlying concept is that there exists a dose-dependent continuum of toxic effects, in which clinically obvious effects have their subclinical counterparts.21A pandemic of subclinical neurotoxicity is therefore likely to be “silent”, i.e. not apparent from standard health statistics.
The concept of subclinical toxicity traces its origins to pioneering studies of lead toxicity in clinically asymptomatic children undertaken by Landrigan and colleagues7and by Needleman and colleagues,8 who showed that children’s exposure to lead could cause decreases in intelligence and alteration of behaviour even in the absence of clinically visible symptoms of lead toxicity. The subclinical toxicity of lead in children has subsequently been confirmed in prospective epidemiologic studies.22,23
Parallel findings have been reported on some other industrial chemicals, but their number is comparatively small. Approximately 80,000 chemicals are registered for commercial use with the US Environmental Protection Agency, and 62,000 were in use when the Toxic Substances Control Act was enacted in the USA 1977.24 The situation is parallel in the EU, where 100,000 chemicals were registered in 1981.25 The full extent to which these chemicals contribute to causation of NDDs and subclinical neurotoxicity is still virtually unknown.
Identification of neurotoxic chemicals
Studies in experimental animals support the notion that a wide range of industrial chemicals may cause developmental neurotoxicity at low doses that are not harmful to adults.26,27 Such injury appears to result in permanent changes in brain function that may become detectable only when the animal reaches maturity and the particular function is expressed. Because developmental neurotoxicity may not be apparent from routine toxicology tests,28 identification of neurotoxic chemicals often rests on clinical and epidemiological data.
To identify environmental chemicals that are toxic to the human brain, we searched the Hazardous Substances Data Bank of the National Library of Medicine, where substance identities are linked to adverse effects in humans. We checked the completeness of this list against other data sources and with a previous assessment of the published literature on clinical toxicity.29 Table 1 lists the documented human neurotoxicants according to major classes of industrial chemicals. We have excluded drugs, food additives, microbial toxins, or snake venoms and similar biogenic substances. This list does not include chemicals that have been found to be neurotoxic solely in laboratory animals, for which no systematic listing exists.
Table 1 primarily includes acute toxicants that have caused serious accidents or have been used in suicide attempts, and neurotoxicants mainly causing chronic or delayed disease are likely to be underrepresented.29The largest groups of identified compounds are metals, solvents, and pesticides, while other groups of chemicals are underrepresented. Table 1 should therefore not be regarded as a comprehensive list of all the industrial chemicals that may be neurotoxic in humans.
The substance names listed in Table 1 were used for searches of published literature on developmental neurotoxicity. Based on our critical evaluation, the human developmental neurotoxicants are highlighted in Table 1. Their number is clearly very small. Many additional chemicals that we have not listed are known to be developmental neurotoxicants in laboratory animals,27 but no published information is available on their potential toxicity to human brain development. We therefore undertook a review of each of the documented human developmental neurotoxicants to determine how their toxicity had been first recognised and to learn how recognition of hazard developed and led to control of exposure.
Lead
The neurotoxicity of lead to adults was known already in Roman times, but a report from Australia published 100 years ago was the first description of epidemic lead poisoning in young children; the source of the outbreak was traced to ingestion of lead-based paint by children playing on verandas with peeling paint.30 Further reports of childhood lead poisoning from the USA and Europe followed. Lead poisoning was at that time considered an acute illness, from which a child either recovered or died. Long-term sequelae were first documented in the1940s, when 19 of 20 survivors of acute poisoning were found to have severe learning and behavioural problems.31
Despite those early paediatric warnings, the largely unchecked use of lead in petrol, paints, ceramic glazes, and many other products through much of the twentieth century continued to cause risks of lead poisoning. During the 1970s, widespread subclinical neurobehavioural deficits, including problems in concentration, memory, cognition, and behaviour, were documented in asymptomatic children with elevated blood-lead concentrations.7,8Spurred by recommendations issued by the European Regional Office of the World Health Organization, studies were initiated in many countries; the results corroborated the previous conclusions.32
As a result of the accumulating evidence,many sources of lead exposure came to be controlled, though not all sources, and not in all countries. A 90% reduction in childhood blood lead levels followed the termination of lead additives in petrol.33 Current research on lead neurotoxicity focuses on the shape of the dose-response curve at very low exposure levels that appear to cause surprisingly large functional decrements.22As convincing evidence became recognised,health agencies have serially reduced the permissible concentration of lead in children’s blood. However, the most recent research22 suggests that the current impact of lead exposure on human brain development may be even greater than previously thought.
Methylmercury
Methylmercury, too, was first established as a neurotoxicant in men exposed occupationally.34The developmental toxicity of this organic mercury compound became evident in the 1960s in Minamata, Japan, where an epidemic of spasticity, blindness and profound mental retardation had occurred among infants born to mothers who consumed contaminated fish. After many years of clinical and experimental studies, the source was found to be mercury compounds released into MinamataBay by a plastics plant.35Methylmercury accumulated and reached high levels in locally caught fish. Exposed adults, including mothers of poisoned children, were less seriously affected, if at all.36Similar outbreaks of profound NDDs in the infants of virtually unaffected mothers have occurred following maternal consumption during pregnancy of seed grain treated with methylmercury fungicides.37,38Studies of a serious poisoning incident in Iraq established a crude dose-response relationship between maternal hair-mercury concentrations and risk of neurological abnormalities in their children.39
Recent studies have focused on prenatal exposures to lower levels of methylmercury. They have examined populations with a high intake of seafood and freshwater fish with various degrees of methylmercury contamination. Prospective examination of a New Zealand cohort found a three-point decrement in IQ as well as alterations in affect among children born to women with hair mercury levels above 6 µg/g.40 A large prospective study in the Faroe Islands found evidence for dose-related impairments inmemory, attention, language and visuospatial perception.41 Athird prospective cohort study in the Seychelles provided only limited support for prenatal neurotoxicity after adjustment for postnatal exposures.42Several cross-sectional studies found significant associations between methylmercury exposure and neurobehavioral impairment in young children.43The US National Academy of Sciences reviewed these studies and concluded that strong evidence exists for the foetal neurotoxicity of methylmercury, even at low exposure levels.44These findings have led food safety authorities to issue dietary advisories, and national and international agencies (with coordination from the United Nations Environment Programme) now seek to control and limit mercury releases to the environment. Substantial reductions have already been achieved in mercury use and release from hospitals and incinerators.45 A related substance, ethylmercury, has been widely used as part of preservatives of vaccines, but a neurotoxic risk has not been documented.46
Arsenic
Ingestion of arsenic-contaminated drinking-water has long been recognised to cause peripheral neuropathy in adults.47 Dramatic documentation of developmental neurotoxicity due to arsenic was reported in 1955 from Japan, where consumption of powdered milk contaminated with arsenic led to over 12,000 poisoning cases and 131 deaths.48A follow-up study of three groups of adolescents born during the time of the milk contamination included one group that was fully breastfed, one that was exposed to the tainted milk product, and one that received other supplements, but no tainted formula.49 Significantly more mentally-retarded individuals were found in the exposed group. Poor school records, emotional disturbances and abnormal or borderline EEG findings were also more common in the exposed group. Because these findings were initially published in Japanese journals not easily available elsewhere,48,49 they have been overlooked, even in the most thorough risk assessments of environmental arsenic exposure.50,51
Arsenic occurs in ground water worldwide, and industrial pollution is widespread. Cross-sectional studies of children at school age showed cognitive deficits associated with drinking water contamination levels52 and increased urinary arsenic concentrations.53 Similar results were obtained in children with arsenic exposure from a smelter.54A possible joint adverse effect on IQ caused by arsenic and manganese exposures was suggested by hair metals concentrations in children living near a hazardous waste site.55Although the recent evidence for subclinical neurodevelopmental neurotoxicity of arsenic is less well developed than for lead and methylmercury, the data are internally consistent and fit with the high-exposure findings from Japan. Still, regulatory action does not emphasise the need to protect the developing brain against this neurotoxicant.50,51