header graphic


How do chemicals end up in my body?

How many chemicals are in me?

Can low doses of chemicals hurt me?

What are the possible health effects of low doses?

Aren't these chemicals well-tested?

What must the chemical industry do?

What should the government do?

What can you do?

These fact sheets are from our print report. Download the full report (in Adobe Acrobat format).

Newsfeed from
Environmental Health News


Can low doses of chemicals hurt me?

Many of the exposures reported here are below the levels thought to be toxic in standard high dose toxicology studies relied upon by industry and regulators. The government’s historic dependence on high dose studies has created an institutional and scientific bias that encourages regulators and industry to assert, with little supporting data, that low doses like those reported here cause no adverse effects.

In this context, the standard response by industry representatives to stories involving chemical contamination is that these “trace” doses are too tiny to cause adverse effects. The science, however, leads to the opposite conclusion. Adverse health effects from low dose exposures are occurring in the population, caused by unavoidable environmental contamination with PCBs, DDT, dioxin, mercury and lead.

Although the chemical industry will assert that low dose exposures to hundreds of chemicals simultaneously is safe, the safety claims are based on a lack of scientific information on the toxicity of low-dose exposures, not on a definitive, scientific proof of safety.

Roots of the myth of low dose safety

Chemical toxicology today falls into two camps: Regulatory toxicology, where scientists, generally in the pay of chemical companies, conduct high dose animal studies under prescribed protocols for the purpose of meeting government requirements; and research toxicology, primarily conducted at independent university and government research centers, where scientists focus on low dose exposures to chemicals that lead to subtle but potentially harmful effects on the body.

Regulatory toxicology targets relatively crude measures of toxicity such as cancer, birth defects, and obvious signs of organ damage. Research toxicology goes beyond this to look at how chemicals can alter the functioning of organ systems that otherwise appear intact.

Many adverse effects are caused by low dose exposures that occur during critical periods of fetal development or infancy but do not manifest until later in life. Often these effects arise from insults that trigger a cascade of effects that alter the proper functioning of organ systems, sexual development, behavior or reproduction. Some examples include alterations of nervous system development that play out as behavioral problems or IQ deficits; or disruption of normal hormonal signaling that results in fertility problems, birth defects of reproductive organs, early puberty, or cancers of the reproductive organs. Alteration of immune system function can also occur, leading to increased susceptibility to illness and disease. Scientists are finding that many chemicals in widespread use today cause these types of effects at doses well below those thought to cause “no effect” in high dose regulatory studies.

The following is a guide through some of the highlights and basic concepts of low dose chemical toxicity. The discussion is presented in greater detail below, but can be summarized as follows:

  • In many cases low doses are toxic. Science has evolved considerably since the 16th century when Paracelsus coined the adage, “The dose makes the poison.” We now know that many other factors, particularly the timing of the dose, genetic variability, and health status of the exposed individual, are equally, if not more important. Low doses of lead, mercury or PCBs at specific days of fetal development or infancy can cause permanent problems that are not manifest during exposure, but only later during childhood, adolescence and beyond (ATSDR 2000, Jacobson and Jacobson 1996, NAS 2000). The same dose in the adult may have little or no effect. Among people breathing the same polluted urban air, only some will develop asthma. A recent study evaluating differences in human response to airborne particles found that the most sensitive members of a population respond to doses 150 to 450-fold lower than median (50th percentile) responders (Hattis, et al. 2001). Factors that contribute to these differences include variations in breathing rates, deposition and elimination of air particles from the respiratory tract, and differences in lung response to the chemicals found in the air.
  • Susceptibility to disease results from a complex interplay between biological factors, such as specific genetic traits, and the environment, which would include exposure to other chemicals (environmental, recreational or therapeutic) and lifestyle factors such as stress levels, nutrition, fitness, and smoking status. Genetic variation contributes to the incidence of diseases induced by industrial contaminants. For example, inherited mutations of the BRCA1 and BRCA2 genes account for five to ten percent of all breast cancers (Tripathy and Benz 1997).
  • For most chemicals science provides little or no basis for assertions that low dose exposures are safe, particularly those that start in utero and continue to old age, as is the case with most of the chemicals identified in this study. This knowledge gap is due to study designs, dictated by food and drug regulations, that in general do not require testing of low doses, and in no cases expose animals to chemicals from conception through old age — a so-called womb to tomb study that most accurately comports with real world human exposures.
  • Most safety claims for trace exposures are based on findings from high dose studies. Such extrapolations do not ensure safety. Low dose safety is not well predicted by the high dose studies required for chemicals used in foods and drugs. High does studies tend to look for readily observable, overt toxic effects like cancer, malformations visible at birth, and organ damage, such as liver toxicity. High dose studies usually involve only adult animals. Low dose studies typically look for less obvious effects that involve impaired function of organ systems in animals exposed during development. Examples include decreased sperm counts, altered hormone levels, impaired immune function, altered growth and development of reproductive organs, and behavioral and intelligence deficits. Low dose studies almost always involve exposures in utero, or early life.
  • Many chemicals produce different or even opposite effects at high and low doses — a phenomenon called biphasic dose response. For example, DES, a potent synthetic estrogen, has been shown to simulate prostate growth at 0.02, 0.2, and 2 mg/kg-day, but inhibit prostate growth at doses of 100 and 200 mg/kg-day (vom Saal, et al. 1997). Perchlorate, a component of rocket fuel and drinking water contaminant, causes changes in the size of certain parts of the brain at 0.01 — 1 mg/kg-day, but not at 30 mg/kg-day (Argus 1998).
  • Current regulatory high dose studies are conducted with purified single chemicals. In the real world, people are exposed to exotic low dose mixtures of several hundred different chemicals. The toxicity of these mixtures is not known, and is not being investigated. In two recent studies scientists dosed laboratory animals with a mixture of 16 organochlorine chemicals, lead, and cadmium, each applied at its individual regulatory “safe” dose, and found that the animals developed impaired immune response and altered function of the thyroid, a gland that is critical for correct brain development (Wade, et al. 2002a, Wade, et al. 2002b).
  • Some people are more sensitive to low dose exposures than others. EPA-funded research has documented a 10,000 fold variability in human response to certain airborne particles (Hattis, et al. 2001). This genetic variability in response explains, in part, why we all breathe the same air, but not all of us have asthma attacks.

Documented low dose effects in people

The scientific evidence for human harm from industrial chemicals and pesticides extends far beyond occupational exposures. Countless studies in the peer-reviewed literature show that adverse health effects from low dose exposures are occurring in the population, caused by unavoidable contamination with PCBs, DDT, dioxin, mercury, lead, and other chemicals. Among the many health effects scientists have linked to chemical exposures in the general population, are premature death, asthma, cancer, chronic bronchitis, permanent decrements in IQ and declines in other measures of brain function, premature birth, respiratory tract infection, heart disease, and permanent decrements in lung capacity (EPA 1996, EPA 2000, Gauderman, et al. 2002, Jacobson and Jacobson 2002, Jacobson, et al. 2002, Kopp, et al. 2000, Longnecker, et al. 2001, NAS 2000, NTP 2002, Pope, et al. 2002, Salonen, et al. 1995, Sydbom, et al. 2001).

PCBs at 9.7 ng/ml (parts per billion or ppb) in maternal serum during fetal development can cause adverse brain development, and attention and IQ deficits that appear to be permanent (Jacobson and Jacobson 1996). Notably, it was the maternal PCB levels and not the PCB levels in children at 4 and 11 years of age (by which time child PCB levels had decreased substantially) that was associated with IQ deficit — underscoring the limitations of studies that try to correlate current body burdens with adverse health outcome in the absence of measuring in utero exposures.

Dioxin at 80 parts per trillion in paternal — but not maternal — serum causes a significant change in the sex ratio of children (Mocarelli, et al. 1996, Mocarelli, et al. 2000). At this tiny dose, men father nearly twice as many girls as boys. Eighty parts per trillion is equivalent to one drop of dioxin in a seven mile long string of bathtubs (7,400 bathtubs).

Lead above 100 parts per billion in the blood of a two year old can cause learning deficits, behavioral problems and a significant decrease in IQ in adolescence and adulthood (CDC 1997). The same dose has no effect on adults. One hundred ppb is the equivalent of 1 drop of water in 6 bathtubs. A 5/1,000ths ounce chip of lead paint can put a child in the emergency room with lead poisoning [Calculated based on (CDC 1997, EPA 1998a)].

Methylmercury causes measurable delays in brain function in children exposed to levels corresponding to 58 parts per billion in maternal blood (NAS 2000).

DDE above 15 ppb in maternal blood is associated with preterm birth, and low birth weight, with weight corrected for gestational age (Longnecker, et al. 2001). DDE is a metabolite of DDT. Using the associations derived from tests of archived samples from a pool of 42,000 women, researchers estimated that DDT exposures could have accounted for up to 15 percent of infant deaths during the 1960s. Low birth weight like that linked to DDE is increasingly recognized as a risk factor for insulin resistance or Type II diabetes, high blood pressure, and cardiovascular disease later in life (Godfrey and Barker 2001, Hales and Barker 2001). Even if these lower birth weight babies “catch up” later, the damage may have already been done. A substantial number of studies have found that low birth weight followed by an accelerated growth rate during childhood is a significant risk factor for high blood pressure, stroke, insulin resistance and glucose intolerance (Eriksson, et al. 2000a, Eriksson, et al. 2002, Eriksson, et al. 2000b, Eriksson, et al. 1999, Eriksson and Forsen 2002, Forsen, et al. 2000, Ong and Dunger 2002, Stettler, et al. 2002).

Chlorpyrifos (dursban) above 8 pg/g (parts per trillion) in the blood of non-smoking women was strongly associated with decreased birth weight and body length in babies of African American women in New York City (Perera, et al. 2003). In the same study, increased air exposure to PAH’s was correlated with decreased birth weight — an effect that was independent from the chlorpyrifos finding — and decreased head circumference. The babies of women exposed to the highest PAH levels had a 10% reduction in body weight (Perera, et al. 2003).

Byproducts of tap water chlorination were linked to statistically significant increases in birth defects in New Jersey at 40 parts per billion in water, and miscarriages in California at 75 parts per billion (Bove, et al. 1995, Waller, et al. 1998).

Perchlorate in drinking water at levels as low as one to two ppb, (0.2 to 0.4 ug/kg/day) is associated with altered thyroid hormone levels in infants (Schwartz 2001). Perchlorate is a component of rocket fuel that also was used in the 1960s as a drug to regulate thyroid hormone activity. Adequate levels of thyroid hormone are critical for normal brain development (Gruters, et al. 2002, Zoeller, et al. 2002).

Low dose effects in animals

The dangers of nearly every chemical banned or restricted in the US were first identified in laboratory animals or wildlife. Animals are strong predictors of hazards to human health — a premise that applies to “all of experimental biology and medicine” (Klassen 1996). For instance, the vast majority of known and reasonably anticipated human carcinogens cause cancer in laboratory animals. (NTP 2002).

As researchers continue to study the effects of exposure to low levels of contaminants, more effects are observed — especially in developmentally exposed organisms. In the natural environment, low dose effects are often observed in aquatic species, such as fish and frogs. This finding has prompted chemical industry representatives to belittle these results as irrelevant to human exposures. The truth, however, is not that simple. Even though the specific effects may differ between humans and wildlife, the general toxicity is often quite similar. For example, if a contaminant causes reproductive effects in fish - such as production of a hormone that humans don’t possess or other effect not caused in humans - the chemical is also likely to affect reproduction in mammals. DES (a potent estrogen) increases levels of vitellogenin — a hormone that humans do not produce — in fish. Vitellogenin is an indicator of estrogenic activity (Folmar, et al. 2002). In mammals, DES causes increased growth of uterine cells, again an indicator of estrogenic activity (EPA 1998c). In either case, DES causes adverse reproductive effects in fish, mammals and humans, although the specific endpoint may differ.

Bisphenol A. A number of low dose studies have focused on effects of bisphenol A, a building block of polycarbonate plastics that is used in dental sealants and to line includely all aluminum and steel cans, among many other uses. The seminal study, by Nagel et al (1997), found increased prostate weight in male mice exposed as fetuses to 2 mg/kg/d. In subsequent studies, scientists have now linked low dose bisphenol A exposures to altered development of the mammary gland (25 mg/kg/d and 100 mg/kg) (Colerangle and Roy 1997, Markey, et al. 2001), vagina (100 mg/kg/d) (Schonfelder, et al. 2002a) and prostate (2 - 50 mg/kg/d) (Gupta 2000, Nagel, et al. 1997, Ramos, et al. 2001); earlier onset of puberty in female mice (2.4 and 20 mg/kg/d) (Honma, et al. 2002, Howdeshell, et al. 1999); effects on behavior (2 to 40 mg/kg/d) (Adriani, et al. in press 2003, Dessi-Fulgheri, et al. 2002, Facciolo, et al. 2002, Farabollini, et al. 1999, Kawai, et al. in press, Palanza, et al. 2002) and decreased sperm production (20 mg/kg/d) (Sakaue, et al. 2001, vom Saal, et al. 1998). Scientists found increased rates of embryonic development at 1 nM (0.23 ppb) ((Takai, et al. 2000a, Takai, et al. 2000b).

Infants ingest bisphenol A in formula at an estimated daily rate of 1.6 mg/kg-day (SCF 2002), giving little safety margin from the doses that cause effects in animal studies (doses as low as 2 ug/kg/d).

Human fetal plasma BPA levels were recently reported at between 0.2 to 9.2 ng/ml (ppb) (Schonfelder, et al. 2002b). The median BPA level in this study (2.3 ng/ml (ppb)) is consistent with a median of 2.2 ng/ml (ppb) reported in a recent Japanese study (Ikezuki, et al. 2002). Notably, some of the effects cause by BPA in animal studies appear to be increasingly common in some segments of the human population, including early onset of puberty (Herman-Giddens, et al. 1997) and decreased sperm production (Swan, et al. 2000, Toppari, et al. 1996).

Atrazine. Five studies published in the past year have found that exposure to 100 parts per triillion of atrazine in water causes deformities in frogs, including hermaphroditism (individuals with both male and female sex organs), underdeveloped testes, and a decrease in the number of germ cells (sperm and eggs) (Hayes, et al. 2002a, Hayes, et al. 2002b, Hayes, et al. in press, Tavera-Mendoza, et al. 2002a, Tavera-Mendoza, et al. 2002b). Hermaphroditism is extremely rare and was not detected in any unexposed frogs (Hayes, et al. 2002b). Atrazine is the most commonly used weed killer in U.S. agriculture, and is found in the tap water of 10 million people in corn belt states. The level that causes these effects, 100 parts per trillion, is commonly found in corn belt tap water and is 30 times less than the legal maximum contamination limit for atrazine of 3 parts per billion.

Aldicarb. Numerous studies have found that low doses of aldicarb impair immune function at low doses (Dean, et al. 1990a, Dean, et al. 1990b, Hajoui, et al. 1992, Olson, et al. 1987, Selvan, et al. 1989, Shirazi, et al. 1990). Immunologic effects were observed at concentrations al low as 0.1 to 1 ppb (Dean, et al. 1990b, Olson, et al. 1987, Selvan, et al. 1989).

Nonylphenol. an ingredient of certain plastics and a surfactant used in detergents and pesticides produces low dose effects in aquatic organisms (i.e. fish and frogs). Nonylphenol at less than 1 ppb in water produced female specific proteins in male fish (Tabata, et al. 2001), altered reproductive hormone levels (Giesy, et al. 2000, Harris, et al. 2001) and decreased sword length in swordtail fish (Kwak, et al. 2001). Slightly higher concentrations (~ 22 ppb) cause an increase in the number of female-appearing frogs (Kloas, et al. 1999). In frogs, low concentrations of NP (100 nM; ~ 22 ppb) decreased the number and differentiation of neural crest-derived melanocytes (pigment producing cells); and this effect was specific to estrogenic compounds tested (Bevan, et al. 2002). Nonylphenol is one of the most frequently detected contaminants of streams in the U.S. (Kolpin, et al. 2002). It was found in 50% of 139 streams in 30 states (median 0.8; max 40 mg/L or ppb).

DDT at 18 ng/ml (ppb) in the blood of mice caused significant increases in the height and thickness of uterine and vaginal epithelial cells respectively. These changes are considered to be indicators of estrogenic response. Changes in uterine epithelial cell height were also observed at b-HCH of 42 ng/ml (ppb) (Ulrich, et al. 2000).

Trenbolone, a synthetic androgen (male hormone) used in beef production, impaired reproduction of fish at 50 parts per trillion (decreased number of eggs spawned) (Ankley and Touart 2002), and caused the “masculinization” of female fish at doses as low as 5 mg/L (5 ppb) — the female minnows grew characteristic male spikes on the tops of their heads.

Regulatory requirements don’t include low dose studies.

For most chemicals scientists have not studied the effects of low dose exposures, particularly in utero. Most available toxicity data comes from high dose studies on adult animals. The dearth of low dose data does not stop industry representatives from issuing blanket assurances about the safety of low dose exposures. When “experts” assert that low doses cause no effects in animals, it is almost always because they haven’t looked.

Most experiments designed to identify low dose toxicity differ substantially from those used in standard high dose studies required for food additives and drugs. These low dose studies are sometimes referred to as non-guideline studies, because they involve investigations that extend beyond the narrow limits of agency study guidelines or protocols.

Research scientists (primarily academic and government researchers) have more flexibility to develop innovative study designs and investigate more sophisticated and subtle indices of toxicity. In contrast, industry scientists typically conduct “guideline” studies that fulfill minimal regulatory requirements, which are often based on decades old science and relatively crude endpoints.

High dose regulatory studies do not look for the outcomes that are most likely to arise from low dose exposures. For example, high dose study designs look for overt easily observable effects, like cancer, gross birth defects, acute poisoning, or overt organ damage. Low dose research studies typically examine functional deficits, where apparently healthy organs or systems, do not function properly. Outcomes are measured as altered growth and development of reproductive organs, behavioral changes, abnormal immune function and changes in hormone levels.

The vast majority of “low dose” studies involve follow-up of developmentally-exposed animals (or humans) in ways not addressed by regulatory toxicology studies. For example, there is no regulatory study that follows animals exposed in utero to an age corresponding to old age — referred to as a “womb to tomb” study. Instead, animals exposed in utero are followed, at a maximum, until they are young adults (~ 4 to 5 months). This makes it impossible to address questions such as whether in-utero or early life exposure to industrial chemicals or pesticides can predispose an individual to cancer, degenerative nervous system disorders, diabetes, or other diseases more prevalent at the end of life for the vast majority of chemicals. In general, research studies follow developmentally exposed animals for longer periods of time and study endpoints in greater detail.

Low doses studies often reveal toxic effects at levels high dose studies consider safe.

High dose animal studies cannot accurately predict either the safety or hazard of low dose exposures. This is particularly true when high dose studies on adult animals are applied to low dose in-utero or infant exposures. Lead, mercury and PCBs are classic examples, where high dose animal studies on mature animals failed to identify the hazards of low dose fetal and childhood exposure.

In many other cases low dose research found adverse health effects at levels well below the supposed “no effect” level determined in standard high dose regulatory guideline studies (Figure 3).

  • Atrazine was presumed by pesticide manufacturers and the EPA to cause no effects below 3 ppb (EPA 2002). Subsequent work by Tyrone Hayes shows that atrazine produces frogs with both male and female sex organs at levels 30 times lower than this (Hayes, et al. 2002b).
  • Methoxychlor, an insecticide and chemical relative of DDT, was presumed to cause no effects for the fetus at doses below 5 mg/kg/d (IRIS - Integrated Risk Information System 2003b). Subsequent low dose research found that methoxychlor causes significant changes in prostate size at a dose 250 times lower in animals exposed in utero (Welshons, et al. 1999).
  • Bisphenol A is assumed to cause no harmful effects below a dose of 5 mg/kg/d, according to a recent risk assessment conducted by European Commission Scientific Committee on Food (SCF 2002). Yet significant effects on reproductive organ size and development have been found repeatedly at levels up to 1000 times lower (Colerangle and Roy 1997, Gupta 2000, Howdeshell, et al. 1999, Markey, et al. 2001, Nagel, et al. 1997, Ramos, et al. 2001, Sakaue, et al. 2001, Schonfelder, et al. 2002a, vom Saal, et al. 1998).
  • Methylmercury is assumed to cause no harmful effects below a concentration of 11 mg/kg in hair according to the Environmental Protection Agency (NAS 2000). Yet researchers in the Netherlands found a doubling in the risk of heart attacks and death from coronary heart disease at methylmercury levels of 2 mg/kg in hair, or about one fifth of assumed safe levels (Salonen, et al. 1995). Increased diastolic and systolic blood pressure and decreased heart rate variability in developmentally exposed children have also been observed at doses below the EPA no effect level (NAS 2000, Sorensen, et al. 1999).
  • Dibutyl phthalate (DBP) was presumed by the EPA to cause no harmful effects in animals below 125 mg/kg/d based on a 1953 study (IRIS 2003a). More recent studies have shown that DBP causes male reproductive toxicity at 100 mg/kg/d, including delayed puberty, cellular changes in the testis and retained nipples (CERHR 2000, Mylchreest, et al. 1999, Mylchreest, et al. 2000). Decreased numbers of live pups have been observed at an even lower dose of 52 mg/kg/d (CERHR 2000, Wine, et al. 1997).

Many chemicals produce different or even opposite effects at high and low doses — a phenomenon called biphasic dose response.

Chemicals that produce a biphasic dose response are relatively common, and these responses are observed for a variety of effects. The prevalence of these types of effects underscores how wrong one could be by assuming that high dose studies accurately predict low dose toxicity.

  • Perchlorate, a component of rocket fuel and drinking water contaminant, causes changes in the size of certain parts of the brain at 0.01 — 1 mg/kg-day, but not at 30 mg/kg-d (Argus 1998). We know that perchlorate causes these effects at lower doses because it is a relatively well-studied chemical by virtue of its use as a pharmaceutical. In contrast, most environmental contaminants would not be assessed for effects on thyroid hormone or brain structure at all.
  • Bisphenol A (BPA), an estrogenic endocrine disruptor commonly found in plastics used in dental sealants and as liners in most aluminum and steel cans, can produce opposite effects at low and high doses. BPA increases the developmental rate of embryonic cells at 1 nM (0.229 ppb), while concentrations 100,000 fold higher (100 mM or 22829 ppb) will decrease developmental rate (Takai, et al. 2000a, Takai, et al. 2000b). In prostate cancer cells, BPA will increase cell proliferation at concentrations 100 times less than the levels that inhibit cell growth (1 vs 100 nM or 0.229 vs 22.9 ppb)(Wetherill, et al. 2002).
  • Atrazine produced more pronounced hermaphroditism and testicular toxicity in frogs at 0.1 ppb than at 25 ppb (Hayes, et al. in press).
  • Pyrethroid insecticides induce hyperactivity in rats at doses up to 0.7 mg/kg but no hyperactivity at a dose 60 times higher (42 mg/kg) (Eriksson, et al. 1991).
  • DES, a potent synthetic estrogen, has been shown to cause stimulatory low dose effects on the weights of the prostate (0.02, 0.2, and 2 mg/kg-day) and uterus (0.1 mg/kg-day), but inhibit growth at higher doses, 200 mg/kg-day and 100 mg/kg-day, respectively (Alworth, et al. 2002, vom Saal, et al. 1997).


While most industry representatives dismiss low dose adverse effects, some embrace a concept called hormesis -— a low dose biphasic dose response where low dose effects are beneficial and high doses are toxic. The concept of hormesis is easily conceptualized with vitamins; low doses of many vitamins are beneficial, while high doses can cause adverse effects including kidney toxicity (vitamin D), gastrointestinal upset (vitamins A and D), headaches (vitamin A), increased susceptibility to hemorrhage (vitamin E) and general sense of fatigue (vitamins A and E) (Merck & Co. Inc. 2002).

Although there is considerable scientific support for hormesis with respect to vitamins and minerals, some in the chemical industry are distorting the concept to argue that the low doses of environmental contaminants may also be “beneficial.” A recent report on hormesis by the Texas Institute for Advancement of Chemical Technology (TIACT), a “non-profit, charitable organizationÉ dedicated to the advancement of chemical technology through an informed public,”,contains the most distorted arguments put forth to date. The authors propose hormesis as a rationale for bringing back into commerce long-banned chlorinated chemicals such as PCBs and DDT.

“The scientific acceptance of hormesis with its possible benefits at low-level exposure could come at no better time than the present when environmentalists and others are calling for bans on more and more chemicals, such as chlorinated hydrocarbons (emphasis added) to prevent low-level exposures. Furthermore, the low-exposure paradigm (emphasis in original) would make it possible for society to enjoy, safely, the benefits of many chemicals that have been banned in the past or could be banned in the future” (page 1 of the executive summary) (TIACT 1998). TIACT is supported by donations from Dow, BASF, Bayer, Shell Chemical Company, and Syngenta.

This extreme view of hormesis is not generally accepted. However, hormesis does help explain low dose effects seen in toxicology studies. Scientists have found that while low doses can stimulate a process in the body, high doses can inhibit the same process. For example, low doses of estrogens will stimulate breast cancer cells to grow (proliferate). High doses of the estrogen can inhibit cell growth — presumably because the high doses can damage the cell to the point of dysfunction or death (Lippman, et al. 1976). Similarly, low doses of pharmacological estrogens stimulate uterine growth in rodents, while high doses — well above therapeutic doses — will cause uterine weight to decrease (Alworth, et al. 2002, Shelby, et al. 1996).

Some people are more sensitive to low dose exposures than others.

People differ in response to the same amount of chemical exposure as a function of their age, differences in metabolic and detoxification pathways, nutritional state, body weight, genetic variability, gender, preexisting conditions, and lifestyle (such as smoking and drinking status). In regulatory toxicology, the default factor used to take these differences into account, referred to as an intraspecies factor, is 10-fold — meaning that the response from one person to another is expected to be no greater than 10 times different.

Chemical response

The assumption of 10-fold variability is not likely realistic when one considers the range of response in the most sensitive populations of people, rather than simple average differences. For example, recent EPA funded research found that some people are 10,000 times more sensitive than the average (median) person to certain forms of airborne particles (Hattis, et al. 2001).


In general, fetuses, infants and children are more sensitive to chemical exposure than adults. One reason is age-related differences in metabolism. A comparison of the half-lives (a measure of how fast a chemical leaves the body) for 45 different pharmaceuticals in neonates and adults found that on average it takes neonates 3 to 9 times longer to get eliminate 1/2 of the administered dose depending on the primary elimination pathway for that chemical (such as CYP or P450, glucuronidation, renal, other non-CYP elimination pathways) (Ginsberg, et al. 2002a).

But averages can mask significant differences. Approximately seven percent (6/85) of 1-week (< 7 days) to 2-month old babies had an elimination half-life more than 10 times longer than the adult average level (Hattis, et al. in press). Only one percent (1/85) of the 1-week to 2-month old infants had a faster half-life than the adult average value (Hattis, et al. in press).

The enzyme paraoxonase (PON) is essential to metabolize toxic breakdown products of organophosphate (OP) compounds, including OP insecticides. People with high PON levels metabolize insecticides faster than people with lower PON levels (Hulla, et al. 1999). Human infants do not begin to produce adult-type levels of PON until they are around 2 years of age (Ecobichon and Stephens 1973), making them potentially more vulnerable to OP exposure.

Similarly, the elderly are also more sensitive to chemical effects, which is why the recommended dosage for many drugs is 25 to 50% of that given to younger adults.

Hexachlorobenzene (HCB), an organochlorine pesticide, is more toxic to the young than to adults. In Turkey during the mid to late-1950s, a fungicide containing ten percent HCB was used to make bread, resulting in an extremely high rate of infant mortality (95%) in breast-fed babies born to mothers who ate the bread. There was no detectable change in mortality for exposed adults (ATSDR 2002b). The infants who died had skin lesions, cardio- respiratory failure, weakness and convulsions. HCB also causes neurotoxicity in adulthood following developmental exposure. Symptoms include a jerkiness of movement like that seen in Parkinson’s disease (ATSDR 2002b). Other effects observed in adults exposed as children include osteoporosis of hand bones, small hands, swelling and spindling of fingers (ATSDR 2002b).

While HCB exposure has not been definitively linked to impaired immune function in humans, exposure to several organochlorines (including HCB) has been associated with increased risk of otitis media (inflammation of the middle ear) in the first year of life (Dewailly, et al. 2000). A German study found that HCB levels were higher in a group of boys undergoing surgery for undescended testicles compared to boys with no testicular abnormalities (Hosie, et al. 2000). More recently, mothers of men with testicular cancer were found to have higher levels of HCB compared to mothers of men without this disease (Hardell, et al. in press).

Genetic differences

Levels of polycyclic aromatic hydrocarbon (PAH)-DNA adducts, a biological marker of PAH exposure, vary up to 24-fold in a population of normal adults, reflecting significant differences in PAH exposure and response (Dickey, et al. 1997). However, in people who lack an important detoxification enzyme, glutathione S-transferase M1 (GSTM1), PAH-DNA adducts vary by 52-fold.

Activity of an enzyme used to metabolize alcohol (as well as the industrial chemicals toluene, vinyl chloride and 2-methoxyethanol), aldehyde dehydrogenase-2 (ALDH2), vary up to 26 fold between susceptible people in Asian populations and the US median. Similarly, activity of another enzyme important in OP detoxification, malaoxonase, varies 7-fold within humans — and this number does not begin to include differences between adult and children (Sams and Mason 1999). These numbers exceed the default factor of 3.2 fold used to account for pharmacokinetic variability in risk assessment (Ginsberg, et al. 2002b).

Progress in government’s efforts to gather low dose study data.

In 1996 EPA convened an expert committee to develop animal testing protocols for low dose studies, to be conducted for a broad range of industrial chemicals that are suspects for low dose effects. Although the original committee and its successor have met regularly for six years now, they have yet to finalize a single testing protocol. One particular protocol still in draft form is a standard uterine growth test used since the 1930s to flag chemicals that could impair reproduction and development.

The committee’s drafts leave out some critical indicators, like tests for brain function in studies of chemicals that suppress thyroid hormones key to brain growth and development. One of the industrial chemicals known to disrupt thyroid function and potentially impair fetal and infant brain development is a rocket fuel ingredient called perchlorate that contaminates an estimated ten percent of the public water supplies in California and that scientists believe crosses the placenta and passes from mother to infant in breast milk (EWG 2000).