Aluminum salts final content: chapter 2.4

2. Summary of information critical to assessment of toxic under CEPA 1999

2.4 Effects characterization

2.4.1 Ecotoxicology

Below, a brief summary of effects data for the most sensitive aquatic and terrestrial organisms is presented. More extensive descriptions of environmental effects are provided in several reviews (for example, Agency for Toxic Substances and Disease Registry (ATSDR) 2006; Bélanger et al. 1999; Roy 1999a).

When aluminum salts are added to water, they hydrolyse, and monomeric aluminum can be formed in the dissolved fraction. It is the monomeric aluminum, and not the salts, that can adversely affect organisms (Driscoll et al. 1980; Parker et al. 1989; Baker et al. 1990). The following summary focuses, therefore, on the effects of the dissolved (particularly monomeric) forms of aluminum that are produced when aluminum salts dissociate.

2.4.1.1 Aquatic organisms

Most of the research on the impact of aluminum on aquatic life has been related to the impacts of acid rain. In this report, emphasis was placed on the potential toxic impacts of aluminum in waters of neutral or near-neutral potential of hydrogen (pH) as the available information suggests that releases associated with the three aluminum salts being assessed occur primarily into waters of circumneutral pH (Roy 1999b; Germain et al., 2000). As described below, because of this consideration, the most relevant effects data identified were for fish. This assessment report does not provide a detailed examination of potential effects from exposure to polymeric aluminum, as polymeric aluminum is most likely to form, and to cause toxicity, during the neutralization of acidic aluminum-rich waters and this is unlikely to occur in the release scenarios considered in this assessment (Roy 1999b).

The gills are the primary target organ for aluminum in fish (Dussault et al. 2001). Aluminum binds to the gill surface, causing swelling and fusion of the lamellae and increased diffusion distance for gas exchange (Karlsson-Norrgren et al. 1986; Tietge et al. 1988). The resulting damage leads to loss of membrane permeability, reduced ion uptake, loss of plasma ions, and changes in blood parameters relating to respiration. Fish death may result from ionoregulatory or respiratory failure, or a combination of both, depending upon the pH of the water and concentration of waterborne aluminum (Neville 1985; Booth et al. 1988; Gensemer and Playle 1999). Ionoregulatory disturbances prevail at lower pH (e.g., below 4.5) and relate to decreased levels of plasma Na+ and Cl¯ ions (Neville 1985; Gensemer and Playle 1999). At pH levels above 5.5, binding of the positively charged aluminum species to negatively charged sites on the gill surface, with subsequent aluminum polymerization, leads to mucous secretion, clogging of the interlamellar spaces and hypoxia (Neville 1985; Poléo 1995; Poléo et al. 1995; Gensemer and Playle 1999).

Aluminum exposure may also disrupt ionic balance and osmoregulation in aquatic invertebrates (Otto and Svensson 1983). Reduced Na+ and/or Ca2+ uptake in response to aluminum exposure have been documented in crayfish (Appleberg 1985; Malley and Chang 1985), mayfly nymphs (Herrmann 1987) and the water boatman, Corixa sp. (Witters et al. 1984). Aluminum reduced Na+ influx and, to a lesser extent, increased outflux, in Daphnia magna, thereby impairing osmoregulation (Havas and Likens 1985). Aluminum may disrupt the respiratory organs of some invertebrates, such as the anal papillae of the phantom midge, Chaoborus sp. (Havas 1986). Respiratory effects can occur when acidic waters are rapidly neutralized, such as when an acidic tributary enters a larger, neutral receiving stream, leading to the formation of mononuclear and polynuclear aluminum species from the dissolved ion (Gensemer and Playle 1999). These species may bind to or precipitate onto the bodies of invertebrates, creating a physical barrier to respiration. Aluminum has been reported to impair reproduction in Daphnia magna (Beisinger and Christensen 1972), although recent work with Daphnia pulex suggests that adaptive strategies which heighten survivorship and fecundity may occur following long-term exposure to sublethal levels (Wold et al. 2005). Hall et al. (1985) reported that aluminum may reduce the surface tension of water, affecting egg deposition, emergence, feeding and mating behaviour of some stream invertebrates.

2.4.1.1.1 Pelagic

Water pH is known to have a significant effect on the toxicity of dissolved aluminum. Under acidic conditions, aluminum is most toxic in the pH range 5.0-5.5. At more acidic pH, its toxicity decreases, while at still lower pH, aluminum can offer transitory protection against the toxicity of H+ (Muniz and Leivestad 1980; Baker 1982; van Coillie et al. 1983; Roy and Campbell 1995). Elevated concentrations of the cations Ca2+ and Mg2+ reduce the toxicity of metals (Pagenkopf 1983; Campbell 1995), yet there are relatively few results examining the effects of elevated calcium on aluminum toxicity. In fish exposed to aluminum at low pH, elevated calcium has been shown to improve survival (Booth et al. 1988; Mount et al. 1988; Sadler and Lynam 1988), reduce losses of plasma ions (Brown 1981; Sadler and Lynam 1988; McDonald et al. 1989) and reduce accumulation of aluminum on gills (Wood et al. 1988a,b). However, Duis and Oberemm (2001) reported low hatching success and high embryo mortality in vendace, Coregonus albula, exposed to high aluminum concentrations of 2.1 and 2.4 mg/L at low pH (4.75, 5.00) and in the presence of 111 to 117 mg/L calcium. Increasing calcium concentrations to 233 to 256 mg/L had no influence on hatching and survival percentages, suggesting that the toxic effect of high aluminum levels can exceed the protective effect of high calcium.

The toxicity of dissolved aluminum is reduced in the presence of inorganic ligands, such as fluorides, sulphates and silicates, as well as organic ligands, such as fulvic and humic acids (Roy 1999a). It is well established that dissolved organic matter (DOM) in particular influences the speciation and absorption of aluminum. In laboratory studies with fish, the toxicity of aluminum was reduced in the presence of organic acids, such as citric acid (Driscoll et al. 1980; Baker 1982), salicylic or oxalic acid (Peterson et al. 1989), humic acid (van Coillie et al. 1983; Parkhurst et al. 1990; Peuranen et al. 2002) and fulvic acid (Neville 1985; Lydersen et al. 1990a; Witters et al. 1990; Roy and Campbell 1997). In laboratory studies with amphibians (frog eggs and tadpoles), LC50s for aluminum increased (toxicity was reduced) in the presence of DOM. However, in the field, the effects of DOM in attenuating aluminum toxicity are difficult to separate from the influences of pH and aluminum concentration (Clark and Hall 1985; Freda 1991).

Most aquatic toxicity studies involving aluminum have been conducted under conditions of low pH, and a number of these accounted for the solubility of the metal in the experimental design. The general conclusion of these studies is that aluminum toxicity is related to the concentration of dissolved inorganic monomeric aluminum (Roy 1999a).

At pH < 6.0, fish, the salmonids in particular, are among the most sensitive organisms to dissolved aluminum. In soft acidic waters, the LC50 can be as low as 54 µg/L (for Atlantic salmon at pH 5.2), while in chronic studies, a lowest-observed-effect concentration (LOEC) of 27 µg/L was determined for growth (for brown trout [Salmo trutta] at pH 5.0). Some species of algae show a comparable sensitivity. Parent and Campbell (1994) determined a LOEC of 150 µg/L (as inorganic monomeric aluminum) at pH 5.0 with the alga Chlorella pyrenoidosa. While many invertebrates tolerate elevated levels of aluminum, Havens (1990) found that exposures to 200 µg Al/L at pH 5.0 were extremely toxic to Daphnia galeata mendotae and Daphnia retrocurva. France and Stokes (1987) concluded that stress from aluminum exposure was secondary to the stress of low-pH exposure for survival of Hyalella azteca. Results of other studies also suggest that invertebrates are more sensitive to low pH than to aluminum. Amphibians show a similar sensitivity. Freda (1991) summarized her work by concluding that aluminum can be lethal to amphibians that inhabit soft acidic (pH 4 to 5) waters if concentrations exceed 200 µg inorganic Al/L.

At pH 6.0 to 6.5, there are few studies that provide effects estimates in terms of inorganic monomeric aluminum. At pH 6.0, a LOEC of 8 µg/L (inorganic monomeric aluminum) for growth of the alga C. pyrenoidosa can be estimated from the data of Parent and Campbell (1994). Growth of the alga was reduced at this single exposure concentration in media without phosphate. This LOEC is, however, well within the likely range of natural concentrations of inorganic monomeric aluminum in surface water. In comparison, Neville (1985) observed that 75 µg Al/L (as inorganic monomeric aluminum) caused physiological distress to rainbow trout (Oncorhynchus mykiss) at pH 6.1 but not at pH 6.5.

At pH 6.5 to 8.0, there are few effects data available. At neutral or near-neutral pH, aluminum has a tendency to precipitate, and the chemistry of these solutions is difficult to control. While the toxicity of alum in neutral-pH waters has been the subject of many studies, the results are unreliable, due to extreme variation between replicates of the same exposure concentration and between duplicate experiments (Lamb and Bailey 1981; Dave 1985; George et al. 1995; Mackie and Kilgour 1995). However, a no-observed-effect concentration (NOEC) for respiratory activity at pH 6.5 is provided by the results of the study by Neville (1985), who found that rainbow trout tolerated 75 µg Al/L (as inorganic monomeric aluminum) during exposures at this pH. Wold et al. (2005) reported a LOEC of 0.05 mg/L Al for reduced survival and reproduction in Daphnia pulex exposed for 21 days to concentrations ranging from 0.05 to 0.50 mg Al/L (nominal) as aluminum sulphate. The test water was maintained at a pH of 7 ± 1, suggesting that the observed effects were due to the presence of aluminum hydroxide rather than the dissolved inorganic monomeric aluminum that is usually associated with toxicity. In addition, the study reported that clonal populations of D. pulex derived from a lake with ongoing alum treatment showed higher age-specific survivorship, higher fecundity and faster growth rates than those collected from waters having less recent or no prior alum exposure. The researchers hypothesized that Daphnia may be capable of exhibiting adaptive strategies that heighten survivorship and fecundity when exposed to sublethal chemical stresses.

Gopalakrishnan et al. (2007) reported a lowest 24-hour EC50 value of 0.210 mg/L for development of the trochophore larva in the marine polychaete, Hydroides elegans. The study was conducted at a pH of 8.1 and aluminum concentrations (measured using atomic absorption spectrophotometry) were well maintained within 2% to 15% of nominal values. Differential sensitivities were observed during embryogenesis and larval development, with lowest toxicity evident at the stage of the fertilization membrane and successively higher toxicity at the blastula and trochophore stages, respectively.

At pH > 8.0, LOECs for survival of rainbow trout are ≥ 1.5 mg/L as total aluminum (Freeman and Everhart 1971). In a more recent study, Gundersen et al. (1994) reported LC50s for exposures of rainbow trout in the pH range 8.0-8.6. The LC50s at all pHs were approximately the same value, ~ 0.6 mg/L (range: 0.36-0.79 mg/L) as dissolved aluminum (filterable through a 0.4-µm filter), and were similar in both acute (96-hour) and longer-term (16-day) exposures at hardness levels ranging from 20 to 100 mg/L (as calcium carbonate). A NOEC for mortality of 0.06 mg dissolved Al/L can be derived from data given for one of the 16-day exposures conducted at 20 mg/L hardness and pH 8.0. Although these concentrations were measured as dissolved aluminum, it is probable that the monomeric aluminate ion, AlOH4- , predominated at this pH.

In contrast, Poléo and Hytterød (2003) reported that juvenile Atlantic salmon, Salmo salar, exposed under alkaline (pH 9.5) conditions to concentrations of around 0.35 mg/L (predominantly aluminate ion) showed no acute toxicity effects. The researchers noted that the aluminum concentrations used in their study were lower than those of Freeman and Everhart (1971) and Gundersen et al. (1994), and hypothesized that more environmentally relevant concentrations of aluminum do not have any acute effect on salmonids under alkaline conditions, while very high concentrations of aluminum might have. While no acute effects were observed, physiological responses in the form of elevated blood glucose and hematocrit levels and a decrease in plasma Cl-, were evident after a three-week exposure period and were considered indicative of a stress response in the fish. The authors concluded that the combination of high pH and aluminum may impose some stress but this is unlikely to represent a serious problem unless the exposure continues for a long period of time. High alkalinity conditions such as those used in the study can occur in water bodies during periods of intense photosynthetic activity in the summer months. At these times, concentrations of aluminum present in the water would also be expected to rise as the solubility of the substance increases over that at lower pH.

While toxicity is most commonly associated with inorganic monomeric aluminum species, there is evidence that aluminum undergoing transition from one species to another is also bioavailable and can exert adverse effects on organisms. Such transition conditions can occur in mixing zones, for example, when acidic waters enter a larger, more neutral receiving system or during the liming of acidic waters. Berkowitz et al. (2005) found that the addition of alum to lake water samples (pH 8.22 to 9.08) resulted in a rapid initial decrease in pH and alkalinity followed by a gradual recovery in pH over several weeks. Dissolved Al concentrations increased following treatment, and then decreased after 150 days. Soucek (2006) determined that freshly neutralized aluminum (i.e., aluminum in transition from ionic species in acidic waters to polymers or precipitating hydroxides after a rapid pH increase) impaired oxygen consumption in Daphnia magna and the perlid stoneflies, Perlesta lagoi and Acroneuria abnormis (lowest LOEC for the study 0.5 mg/L, which was also the lowest concentration tested). Alexopoulos et al. (2003) reported that freshly neutralized aluminum at a concentration of 0.5 mg/L associated specifically with the gills of the freshwater crayfish, Pacifastacus leniusculus, creating a physical barrier during precipitation that resulted in impaired respiration and asphyxiation. Particulate aluminum has been shown to decrease filter feeding in the freshwater bivalve, Anodonta cygnea, presumably as an avoidance response to the toxicant (Kádár et al. 2002). Poléo and Hytterød (2003) examined toxicity under steady-state (pH retained at 9.5) and non-steady state (pH lowered from 9.5 to 7.5) conditions in order to evaluate the possible impact of transient aluminum chemistry on Atlantic salmon, Salmo salar. No increase in toxicity occurred under the non-steady state conditions (where aluminum solubility was lowered as the pH decreased) and the physiological disturbances observed at high pH were mitigated. The results contrasted with those obtained in studies where aluminum solubility was lowered by raising the pH of aluminum-rich water. In these cases, toxicity to fish increased as the solubility of aluminum was decreased and aluminum precipitated onto the gills (for example, Poléo et al. 1994; Poléo and Bjerkely 2000).

Verbost et al. (1995) reported enhanced toxicity in a mixing zone of acid river water containing aluminum (pH 5.1, aluminum 345 µg/L) with neutral lake water (pH 7.0, aluminum 73 µg/L). The resulting water (pH of 6.4, aluminum 235 µg/L) was expected to have low toxicity; however, the freshly mixed water was highly toxic to brown trout, Salmo trutta, with necrosis and apoptosis of the gills evident in exposed fish. A clear gradient in the deleterious effects occurred with increasing distance from the mixing area, with fish furthest from the mixing zone exhibiting only mild effects. The researchers concluded that freshly mixed acid and neutral water contains toxic components during the first seconds to minutes after mixing, and that even short exposure to this toxic mixing zone is detrimental to migrating trout. Farag et al. (2007) hypothesized that colloids formed in mixing zones may contribute to aluminum toxicity in fish by providing a direct route of the metal to the gills.

Finally, in a study done with DWTP sludge from Calgary and Edmonton, Alberta, AEC (1987) concluded that all sludges tested were non-toxic using a microbial test and acutely and subacutely non-toxic to rainbow trout. However, delayed release of first broods and significantly reduced reproduction were reported in the freshwater cladoceran, Ceriodaphnia dubia, exposed for 7 days to 100% aluminum sludge effluent collected from a DWTP in the U.S. (Hall and Hall 1989). The researchers considered that the effects were likely due to the combined effects of reductions in pH and dissolved oxygen concentrations, physical stress due to high levels of suspended solids, and possibly the presence of aqueous aluminum. Aqueous aluminum alone was probably not the factor exerting sub-lethal toxicity in 100% effluent since similar aqueous aluminum concentrations were observed in the 50% effluent where delays and significant reductions in reproduction were not observed. The same study observed significant mortality in fathead minnow, Pimephales promelas, exposed to the 100% effluent, as well as at the lowest test concentration of 6.3%. Mortality in the intervening concentrations was not statistically different from that in the controls. Mortality at 100% effluent was attributed to physical stress resulting from high levels of suspended solids. While a causative agent for the observed mortality at 6.3% could not be identified, the researchers noted that this test concentration had the highest concentration of aqueous aluminum, with measured levels up to 0.43 mg/L as compared with 0.05 to 0.31 mg/L at the other test concentrations. No sublethal impacts were evident in the fish testing.

2.4.1.1.2 Benthic

Alum can be used to treat eutrophic lakes to reduce the amount of phosphorus present in water or prevent its release from sediment. Lamb and Bailey (1981) concluded that a well-planned and controlled alum treatment would not result in significant mortality in benthic insect populations. Connor and Martin (1989) measured no detrimental effects on midge or alderly larvae following treatment of Kezar Lake, New Hampshire, sediment, and long-term effects on benthic invertebrates were minimal. Narf (1990) reported that benthic population diversities and numbers increased or remained the same following lake treatment with alum. Smeltzer (1990) observed a temporary impact on benthos after treatment of Lake Morey, Vermont, with an alum/sodium aluminate mixture. Benthos density, already low in the year prior to treatment, and richness were lower following treatment. However, changes were not significant, the benthic community recovered, and two new chironomids appeared the following year.

The Sludge Disposal Committee examined the impact of alum sludge discharge in aquatic environments and concluded that residue will tend to deposit near the point of discharge if the water velocity is low (Cornwell et al. 1987) and that it could have adverse effects, including development of anaerobic conditions. Roberts and Diaz (1985) related the reduction in phytoplanktonic productivity observed during alum discharge in a tidal stream in Newport News, Virginia, to the reduction in light intensity. Lin et al. (1984) and Lin (1989) found no buildup of sludge in pooled waters in the Vermillion and Mississippi rivers following sedimentation basin cleaning of drinking water treatment plants (DWTPs) in St. Louis, Missouri. There were no significant differences in types and densities of macroinvertebrates in bottom sediments, and even higher density and diversity were found in some sites.

George et al. (1991; 1995) reported that macroinvertebrates located downstream of four DWTPs appeared to be stressed by alum discharges. In the Ohio River, effects seemed temporary and were limited in space. In addition, organisms collected from upstream locations indicated that environmental factors other than the aluminum sludge discharge may also have been affecting the system. A water-sediment microcosm study done with bottom sediment from the receiving rivers over a 72-day period showed significantly lower oligochaete content in bottom sediment treated with alum sludge. Testing with bentonite gave the same results, and the authors concluded that aluminum sludge deposits on sediment may have the potential to detrimentally affect benthic macroinvertebrate populations by limiting their access to oxygen or food and, therefore, the smothering effect from sludge may prove to be more important to aquatic organisms than aluminum content. However, in laboratory testing, filtrates obtained from aluminum sludge were toxic to the freshwater alga, Selenastrum capricornutum, in waters with low pH or a hardness of less than 35 mg/L CaCO³, suggesting that water-soluble constituents from the aluminum sludge may be capable of affecting algal growth. The study recommended that further toxicity testing be conducted to more fully ascertain potential toxic effects, and that aluminum sludge not be discharged into soft surface waters (that is hardness < 50 mg CaCO³/L) or those with a pH of less than 6.

A study has been undertaken to examine the environmental impact of filter backwash and basin cleaning effluents to the Ottawa River from the Britannia and Lemieux Island DWTPs in Ottawa (RMOC 2000; City of Ottawa 2002). In this study, riverine characteristics downstream of the Britannia site were reported to be beneficial for the sampling of benthic invertebrates due to the slow water velocities of a bay environment. Unlike the Britannia site, the Ottawa River in the vicinity of the Lemieux Island DWTP was characterized by strong currents and an absence of natual benthic habitat. To examine the impact of effluents from the Lemieux Island facility, artificial habitat was installed for benthic organisms at both upstream and downstream locations from the discharge site. The results of the sampling showed that species abundance and diversity was depressed at both sites downstream from the effluent discharges in comparison to sampling sites located upstream.  At sites located 150 and 6,000 m upstream from the Britannia DWTP outfall, approximately 160 and 250 organisms were counted, whereas downstream sites located at 0, 300, 500 and 1,500 m (furthest sampling location) had between 3 (at 0 m) and approximately 100 organisms (at 1,500 m) (diversity of organisms not provided for Britannia site). At the artificial sampling sites 30 and 110 m downsteam of the Lemieux Island DWTP, approximately 250 and 1,000 organisms were counted representing 17 and 21 taxa, respectively.  The site located 90 m upstream from the Lemiux discharge had approximately 1,800 organisms representing 24 taxa.

Toxicity of basin sediment from each of the Britannia and Lemieux Island DWTPs was also examined. The studies showed complete mortality of midge larvae (Chironomus riparius) within the 10 day test exposure, while survival of Hyalella azteca (14 day exposure) was not significantly different from that of the control animals. The study could not determine whether the mortality was attributable to the physical characteristics of the sludge (for example, particle size) or the presence of chemical contaninants.  The sludge from the Lemieux Island DWTP was shown to inhibit growth of Hyalella azteca over the 14 day exposure period, but the Britannia DWTP sludge resulted in no observed effect. The study did not suggest why one sludge demonstrated growth effects, but not the other (methodology and experimental conditions were not provided).

Ultimately, the cause of the the depressed levels of organisms downstream in the Ottawa River from Britannia and Lemieux DWTPs was not due to one causal factor, rather may have resulted from a number of attributes including: physical composition of the sediment and its ability to support life; ongoing blanketing of the area due to new discharges; and toxicity of dissolved aluminum leaching out of the sediment into the water column (City of Ottawa 2002).

In studies related to wastewater releases by DWTPs, AEC (1984) reported there is potential for smothering effects on benthic organisms related to settled sludge on sediments following their release to rivers in Alberta. A number of other possible adverse impacts resulting from the discharge of aluminum sludge to receiving waters were identified, including: formation of sludge deposits in quiescent areas of streams; toxic effects on aquatic organisms from other contaminants present in the sludge; periodic high oxygen demand if water treatment plant sludge is discharged in large slugs or if previously deposited sludge is periodically re-suspended due to increased stream velocity; increased aluminum concentrations of downstream water supplies; and aesthetic problems where stream flow, stream turbidity, and/or sludge dilution are low. The researchers concluded that aluminum sludge exhibits a wide range of characteristics which depend on the raw water characteristics (turbidity, etc.) and other factors and, therefore, while numerous suspicions have been expressed regarding the potential for adverse effects resulting from the discharge of alum sludges to receiving waters, there appeared to be a lack of good scientific evidence to substantiate these concerns. Recommendations of the report included the acquisition of baseline data through bioassay testing and other studies, as well as consideration of alternatives to direct stream disposal practices such as reduction of the quantities of alum sludge produced through substitution with other coagulants, discharge at controlled rates to a sanitary sewer, lagooning with natural freeze-thaw dewatering, thickening and dewatering followed by landfilling, and land application.

A subsequent study examining the binding, uptake and toxicity of aluminum sludges from three water treatment systems in Edmonton and Calgary determined that aluminum was effectively bound to sludges within the pH range 4.5 to 10.0, with more than 99.98% of the total aluminum being in the form of sludge (AEC 1987). Sludge collected from the three plants was found to be non-toxic to rainbow trout, Long Evans rats, and the microbial toxicity test system, Microtox.

2.4.1.2 Terrestrial organisms

Research on the effects of aluminum to soil organisms has concentrated largely on screening for aluminum-tolerant strains of root nodulating bacteria and mycorrhizal fungi, due to the importance of these species in improving crop production (Bélanger et al. 1999). In general, toxicity threshold values for bacterial species fall in the range of 0.01 to 0.05 mM (pH 4.5 to 5.5), while those of mycorrhizal fungi range from 0.1 to 20 mM (pH 3.4 to 4.5) when based on hyphal growth inhibition and 30 to 157 mg/kg soil (pH 4.5 to 5.0) when based on reduced spore germination. For soil macroinvertebrates, growth of newly hatched earthworm, Dendrodrilus rubidus, was significantly reduced at 10 mg Al/kg soil (soil pH 4.2 to 4.9; Rundgren and Nilsson 1997), while significantly inhibited growth and cocoon production were reported for the earthworm, Eisenia andrei, at concentrations ranging from 320 to 1000 mg/kg dry soil, with toxicity decreasing as soil pH increased from 3.4 to 7.3 (van Gestel and Hoogerwerf 2001). A more complete examination of potential impacts to soil-dwelling microorganisms, fungi and invertebrates can be found in Bélanger et al. (1999).

The remainder of this section focuses on the effects of aluminum on sensitive plant species. It should be noted, however, that the problem with alum sludge may be associated not only with the direct toxic effects of aluminum on plants, but also with indirect effects related to phosphorus deficiencies (Jonasson 1996; Cox et al. 1997; Quartin et al. 2001). Aluminum's capacity to fix labile phosphorus by forming stable aluminum-phosphorus complexes and hence make it unavailable to plants can be responsible for the observed effects. In addition, toxic substances captured by the floc during water treatment may be available for uptake by soil species and exert adverse effects.

The presence of aluminum in solution, soil solution or soil resulted in a decrease in seedling growth, elongation or branching of roots of hardwood and coniferous species at varying levels (Horst et al. 1990; Bertrand et al. 1995; McCanny et al. 1995; Schier 1996). The most sensitive species was honeylocust (Gleditsia triacanthos) (Thornton et al. 1986a, 1986b). All measures of growth, except root elongation, consistently declined as solution aluminum increased, 0.05 mM or 1.35 mg/L being the critical value for a 50% general decrease (pH = 4.0). Since honeylocust is not an important species in Canadian forests and since the results obtained by Thornton et al. (1986b) contradict the results obtained for this species by other researchers, it was decided that the two next lowest-observed-adverse-effect concentrations (LOAECs) are more relevant. Hybrid poplar (Populus hybrid) (Steiner et al. 1984) and red oak (Quercus rubra) (DeWald et al. 1990) showed a 50% decline in root elongation at an aluminum solution level of 0.11 mM (2.97 mg/L). The most sensitive coniferous species is pitch pine (Pinus rigida) (Cumming and Weinstein 1990). Seedlings inoculated with mycorrhizal fungus, Pisolithus tinctorius, showed increased tolerance to aluminum, whereas non-mycorrhizal seedlings exposed to 0.1 mM (2.7 mg/L) (pH 4.0) aluminum exhibited decreased root and shoot growth.

In an experiment done with scots pine (Pinus sylvestris), Ilvesniemi (1992) found that when nutrition was optimal, pines tolerated high levels of aluminum, but in nutrient-poor solution, their tolerance to aluminum was reduced tenfold. Hutchinson et al. (1986) and McCormick and Steiner (1978) also observed that pines were tolerant of high levels of aluminum in optimal nutrient solution.

Grain crop and forage crop species were also affected by different levels of aluminum (Bélanger et al. 1999). Wheeler et al. (1992) found that two barley (Hordeum vulgare) cultivars and eight common wheat (Triticum aestivum) cultivars were particularly sensitive, growth being decreased by more than 50% at aluminum levels as low as 0.005 mM (0.135 mg/L) (pH 4.5). Wheeler and Dodd (1995) also showed a 50% decline in growth of clover species, Trifolium repens, Trifolium subterraneum and Trifolium pratense, at 0.005 mM (0.135 mg/L) aluminum (pH 4.7). In a solution culture study, Pintro et al. (1996) found that the root elongation rate of maize (Zea maize HS777 genotype) was also negatively affected at an aluminum level of 0.005 mM (0.135 mg/L) (pH 4.4). In a study done on barley, Hammond et al. (1995) found significant amelioration of the toxic effects of aluminum on root and shoot growth when silicon was added to the solution medium. Silicon amelioration of aluminum toxicity in maize has also been reported (Barcelo et al. 1993; Corrales et al. 1997). In the presence of silicon, aluminum uptake seems to be decreased because of the formation of aluminum-silicon complexes, thus leading to a decrease in absorption of aluminum. In addition, complexes formed with organic anions, sulphate and phosphate appear to be non-toxic to plants (Kinraide 1997; Takita et al. 1999; Matsumoto 2000), while the aluminum-hydroxy species was reported to be phytotoxic in early studies (Alva et al. 1986; Wright et al. 1987; Noble et al. 1988a) but not in more recent ones (Kinraide 1997). Complexation with fluoride has been shown to ameliorate the phytoxic effects of aluminum in nutrient solutions (Cameron et al. 1986; Tanaka et al. 1987; MacLean et al. 1992); however, the aluminum-fluoride complex may also become toxic at high concentrations, with toxicity linked to the proportion and concentration of the different types of aluminum-fluoride species present in solution (Kinraide 1997; Stevens et al. 1997). Manoharan et al. (2007) reported severely restricted root growth in barley exposed to fluoride and aluminum in acidic soils (pH 4.25 to 5.48). Toxicity was attributed the activities of AlF2+ and AlF2+ complexes formed in the soil. Fluoride may enter soil through the application of phosphate fertilizers, which usually contain 1% to 4% fluoride as an impurity (Loganathan et al. 2003). Calcium supplementation has also been reported to alleviate aluminum toxicity in barley, possibly by reducing cellular absorption of the metal and enhancing protection through increased activity of antioxidant enzymes (Guo et al. 2006).

Wheeler and Dodd (1995) investigated the effect of aluminum on yield and nutrient uptake of some temperate legumes and forage crops using a low ionic strength solution. The solution aluminum levels at which top yield and root yield of 58 white clover cultivars were reduced by 50% ranged from approximately 0.005 to 0.02 mM (0.135 to 0.540 mg/L) (pH 4.5 to 4.7).

Although inorganic monomeric forms of dissolved aluminum (Al3+, Al(OH)2+ and Al(OH)2+) are believed to be the most bioavailable and responsible for most toxic effects (Alva et al. 1986; Noble et al. 1988b), information on the concentrations of different dissolved aluminum complexes was not reported in many of the effects studies reviewed. For studies indicating particular sensitivity that were carried out in the laboratory in artificial solutions, it is likely that the majority of the aluminum present in these key studies was in inorganic monomeric forms. Considering that solution culture experiments gave lower LOEC values than did sand culture experiments in forest species studies, the effects data reviewed are considered to be conservative estimates of the effects levels for vegetation grown in natural soils.

2.4.2 Experimental mammal studies

The scientific literature concerning the effects of aluminum exposure in experimental mammals is large, including studies with a variety of administration routes (ingestion, inhalation, dermal, intraperitoneal, intravenous, intracisternal). The characterization of effects presented below includes studies of oral, inhalation and dermal administration, with emphasis on the oral exposure studies. This reflects the importance of the oral route in environmental exposures within the general Canadian population, as compared to dermal and inhalation as well as the research emphasis on oral studies within the scientific community. For more detailed discussion of other routes of exposure, the reader may consult the comprehensive reviews cited, in particular Krewski et al. (2007).

Health Canada considers neurotoxicity and reproductive/developmental toxicity as the categories of effects of greatest potential concern for the general population, in light of the evidence from case studies and epidemiological investigations, discussed in section 2.4.3. Recent comprehensive reviews also collectively support this conclusion (InVS-Afssa-Afssaps 2003; ATSDR 2006; JECFA 2006; Krewski et al. 2007; EFSA 2008). Thus, most of the studies presented in this section focus on neurotoxicity or reproductive/developmental toxicity in which aluminum is administered to the experimental animals through diet, drinking water or gavage.

Various aluminum salts, including chloride, nitrate, sulphate, lactate, citrate, maltolate, fluoride and hydroxide have been used in experimental animal studies to investigate the effects of Al3+ absorbed in the bloodstream and distributed to target organs. Aluminum speciation (i.e., the ligands associated with aluminum) and the overall composition of the diet may influence toxicokinetics and consequently the subsequent toxicity of Al3+ (see section 2.3.3.1.1). With respect to absorption, however, no one aluminum salt is representative of the mix of aluminum compounds in the human diet that contribute to the Al3+ reaching the bloodstream. Therefore, for the purpose of characterizing effects of total aluminum, all oral studies were examined, regardless of the aluminum salt administered. Relative bioavailability of particular salts is then considered in the exposure-response analysis of section 3.2.3.

A number of the experimental animal studies are designed to explore the influence of factors that may potentially exacerbate the toxic effects of aluminum (for example, restraint) or provide protection (for example, therapeutic substances such as Gingko). The results reported in this section, however, focus on the differences between aluminum-treated animals and controls, rather than the influence of these other factors.

In most of the studies consulted, there is a lack of data on the aluminum concentration in the base diet. Studies on different brands of commercial laboratory animal chow show that aluminum levels in the chow can be significant relative to the administered doses, and also highly variable between brands and even between different lots of the same brand (ATSDR 2006). Typical levels of 250 to 350 ppm of aluminum in rodent chow (ATSDR 2006) would contribute approximately 13 to 18 mg Al/kg/d in rats and 33 to 46 mg Al/kg/d in mice, on the basis of default reference values for animal intake and body weight proposed in Health Canada (1994). While it may be hypothesized that the absorption of the base diet aluminum may differ from (and be significantly less) than the absorption of the administered aluminum, there are little relevant experimental data on this question (see section 2.3.3). Therefore the lack of data on base diet aluminum in many of the toxicity studies must be considered as a major uncertainty in the overall database, when considering these studies in the exposure-response analysis and risk characterization.

Notwithstanding the importance of quantifying total aluminum exposure in animal studies, in order to provide a qualitative summary of the literature for the purpose of hazard identification, all studies have been evaluated, regardless of whether the base diet aluminum concentration is reported. In the exposure-response analysis (section 3.2.3), however, administered and combined doses are distinguished and the influence of this factor is considered.

The description of the studies in this section is focused on the nature of the effects investigated and observed, rather than the exposure-response relationship. The database is large (138 studies) and the experimental conditions (for example, administered salts and dosing regimen) vary, and in the majority of the studies only one dose was tested. Thus direct comparisons of the dose-effect data may be misleading. While some information on the lowest observed dose at which effects occurred is providedFootnote 14 as well as the highest dose at which no effects were observed, a more detailed discussion of the exposure-response analysis is presented in section 3.2.3. The details of the studies considered in that analysis are summarized in Tables C1 and C2 (Appendix C). Tables summarizing the full dataset are available in the Health Canada Supporting Document, prepared for this draft assessment (Health Canada 2008a).

2.4.2.1 Acute Toxicity
Oral Exposure

The oral LD50 (lethal dose, 50% kill, single administration) for different aluminum salts, as measured in different strains of mice, rats, guinea pigs and rabbits, varies according to the aluminum salt administered as well as according to the experimental animal species. In an early review an LD50 of apparently 6,200 mg Al/kg bw was reported for Al2(SO4)3 and of 3,850 mg Al/kg bw for Al(Cl)3 administered to mice (Sorenson et al. 1974), although it is unclear from the review article if these values refer to the dose in terms of aluminum or the dose in terms of the salts. Sorenson et al. (1974) also reported LD50 values from 260 to 4,280 mg/kg bw for Al(NO3)3·9H2O in two separate studies on rats. The lower value of 260 mg/kg Al(NO3)3·9H2O clearly underestimates the LD50 (that is overestimates the toxicity), as Colomina et al. (2002), Colomina et al. (2005) and Domingo et al. (1996) have shown. These research groups tested administered doses of 50 to 100 mg Al/kg bw/d, equivalent to approximately 700 to 1,400 mg Al(NO3)3·9H2O/kg bw/d, and the effects were limited to alterations in weight gain and subtle neurological effects (see sections 2.4.2.2 to 2.4.2.4 and section 3.3 for more detailed discussion of these studies).

In a study of oral and intraperitoneal administration during 14 days, Llobet et al. (1987) estimated the acute oral toxicity of aluminum chloride, nitrate and sulphate in Sprague-Dawley rats and Swiss mice. Aluminum chloride and nitrate produced acute toxicities of similar magnitude (LD50 of 222 to 370 mg Al/kg) in the mice and rats, whereas the toxicity of aluminum sulphate was considerably lower (LD50 > 730 mg Al/kg in both species).

Inhalation exposure

In Golden Syrian hamsters and New Zealand rabbits exposed over a short duration (four to six hours per day for three to five days at levels of 7 to 200 mg/m³) to aluminum chlorohydrate through inhalation, the effects observed are those typically associated with inhalation of particulate matter, including alveolar wall thickening, increased number of macrophages and increased lung weight (ATSDR 2006). A more detailed discussion of the pulmonary effects in experimental animals of inhalation exposure to aluminum oxide dust and refractory alumina fibres, and aluminum hydroxide is provided by Krewski et al. (2007). The observed responses to various species of aluminum are described as "typical of foreign body reaction", including alveolar proteinosis and wall thickening, and some nodule formation.

Dermal exposure

Dermal effects of aluminum compounds (10% w/v chloride, nitrate, chlorohydrate, sulphate, hydroxide) applied to skin of mice, rabbits and pigs over five-day periods (once per day) include epidermal damage, hyperkeratosis, acanthosis and microabscesses (ATSDR 2006; Krewski et al. 2007).

2.4.2.2 Short-term toxicity (duration of exposure less than 90 days)
Oral exposure

The results of 40 short-term studies in adult mice, rats and rabbits (exposure duration between 3 and 13 weeks) are summarized below. In all the studies considered, aluminum was administered orally in drinking water, in the diet or by gavage. The aluminum salts include lactate, chloride, sulphate, nitrate and hydroxide. In some studies citrate was administered with the aluminum salt in order to enhance absorption.

As discussed in section 2.4.2, many of the short-term studies did not quantify the concentration of aluminum in the base diet. In these cases the value of the actual combined dose is highly uncertain, particularly in the studies where the administered dose was significantly less than the possible baseline dose in the diet (for example, Basu et al. 2000; El-Demerdash 2004; Kaizer et al. 2005; Kaur and Gill 2005, 2006; Jyoti and Sharma 2006; Sparks et al. 2006; Kaur et al. 2006). In three studies (Thorne et al. 1986; Shakoor et al. 2003; Campbell et al. 2004), ambiguities in the reporting of the doses precluded consideration of the dose-response relationship; however the qualitative observations from these studies are included in the following summary of effects.

Neurobehavioural effects in adult rats and mice following oral administration from 21 to 90 days included decreased performance in the rotarod test (Bowdler et al. 1979; Shakoor et al. 2003; Kaur et al. 2006), decreased performance in passive and active avoidance tests (Commissaris et al. 1982; Connor et al. 1988; Connor et al. 1989; Kaur et al. 2006), reduced motor activity (Commissaris et al. 1982; Golub et al. 1989; Shakoor et al. 2003), decreased forelimb and hindlimb grip strength (Oteiza et al. 1993), increased sensitivity to flicker (Bowdler et al. 1979) and air puff startle response (Oteiza et al. 1993), and reduced recovery in neurological function following spinal cord injury (Al Moutaery et al. 2000).

Of the above studies, the lowest administered dose at which effects occurred was observed by Kaur et al. (2006), in which male Wistar rats were administered 10 mg Al/kg bw/d as aluminum lactate for up to 12 weeks, with testing at 0, 4, 8 and 12 weeks. A significant decrease in performance between exposed and control groups was observed at four weeks and became more pronounced following eight weeks of exposure. Decreased performance in memory function tests (passive and active avoidance responses) was also observed in the exposed animals tested at 12 weeks.

In contrast, no alterations in passive or active avoidance test results were reported in aluminum-exposed animals, at doses of 67 mg Al/kg bw/d of aluminum chloride administered by gavage to male Sprague-Dawley rats for 28 days (Bowdler et al. 1979) and 600 mg Al/kg bw/d of aluminum nitrate administered in drinking water for 14 days to male CD mice (Colomina 1999).

Reduced body weight among aluminum-exposed animals was observed by Bataineh et al. (1998), at a dose of 15 mg Al/kg bw/d of aluminum chloride administered to male Sprague-Dawley rats in drinking water for 12 weeks. On the other hand, Colomina et al. (1999) observed a reduction in body weight only at 600 mg Al/kg bw/d of aluminum nitrate, and no effect at 300 mg Al/kg bw/d, in mice administered aluminum via drinking water for 14 days. In other short-term studies, the authors either did not observe this effect, at a dose of100 mg Al/kg bw/d administered in the diet of Swiss Webster mice (Donald et al. 1989; Golub and Germann 1998), or did not report differences in body weight between exposed and control groups.

The most extensive histopathological changes in the short-term studies were reported by Roy et al. (1991a) in which male rats were given doses of 17 to 172 mg Al/kg bw/d as aluminum sulphate via gavage. The concentration of aluminum in the base diet was not quantified. Multifocal neuronal degeneration, abnormal and damaged neurons, and reduced neuronal density were identified in specific brain regions (for example, cerebral cortex, subcortical region and base of brain) at 29 mg Al/kg bw/d. In the liver, Roy et al. (1991a) observed cytoplasmic degeneration in the periphery of the hepatic lobule at all doses. With increasing doses, multifocal degeneration of the entire liver tissue was observed, followed by fibrous tissue proliferation. Kidney effects observed in this study at 22 mg Al/kg bw/d included increased swelling and degeneration of the cortical tubules.

Other histopathological effects reported in different strains of rats include necrosis-like changes in hippocampal CA1 cells and accumulation of synaptic vesicles in presynaptic terminals (Jyoti and Sharma 2006), congestion of cerebral and meningeal blood vessels, multifocal neuronal degeneration, neurofibrillary degeneration and foci of demyelination (El-Rahman 2003), increased membrane fluidity and decreased cholesterol/phospholipid ratio in synaptosomes (Silva et al. 2002), increased number of vacuolated spaces in the matrix of the cerebral cortex (Basu et al. 2000), decreased NADPH-diaphorase positive neurons in the cerebral cortex (Rodella et al. 2001) and increased hippocampal muscarinic receptors (Connor et al. 1988). The lowest administered doses at which such changes occurred were in the studies of Jyoti and Sharma (2006) in which exposed male Wistar rats received a dose of 10 mg Al/kg bw/d of aluminum chloride in drinking water for five weeks, and of Basu et al. (2000), in which male Sprague-Dawley rats received 10 mg Al/kg bw/d of aluminum chloride via gavage for 40 days.

The biochemical changes to the brains of adult rodents resulting from oral administration of aluminum salts for periods of less than 90 days included effects on cholinergic neurotransmission (Kumar 1998; Shakoor et al. 2003; El-Demerdash 2004; Kaizer et al. 2005; Kaur and Gill 2006) as well as changes in the levels of other neurotransmitters and signalling proteins (Flora et al. 1991; Tsunoda and Sharma 1999b; Kumar 2002; El-Rahman 2003; Becaria et al. 2006), alterations in calcium transfer, binding and signalling in the brain (Kaur et al. 2006; Kaur and Gill 2005), evidence of oxidative stress in different regions of the brain (Fraga et al. 1990; Katyal et al. 1997; Abd el-Fattah et al. 1998; El-Demerdash 2004; Nehru and Anand 2005; Becaria et al. 2006; Jyoti and Sharma 2006), changes in ATPase activity (Katyal et al. 1997), alterations to cyclic AMP second messenger systems (Johnson and Jope 1987), increased levels of amyloid precursor protein (Becaria et al. 2006) and increased TNF-µ (alpha tumour necrosis factor) mRNA expression in the brain (Tsunoda and Sharma 1999a; Campbell et al. 2004). The lowest administered dose at which such effects were observed was 10 mg Al/kg bw/d administered to rats as aluminum lactate via gavage or as aluminum chloride via drinking water in Kaur and Gill (2006) and Basu et al. (2000).

Inhalation exposure

The toxicological literature for short-term inhalation exposure studies is limited compared to that for oral exposure. The most recent comprehensive reviews of this literature can be found in ATSDR (2006) and Krewski et al. (2007). The most sensitive and best documented endpoints concern the respiratory system. The observed effects were those commonly associated with particle inhalation exposure (> 7 mg/m³), including a thickening of the alveolar walls, an increase in alveolar macrophages and heterophils, granulomatous nodules and lesions, and increased lung weight (ATSDR 2006).

2.4.2.3 Subchronic and chronic toxicity (exposure duration greater than 90 days, non-cancer endpoints)
Oral exposure

The results of 49 subchronic and chronic toxicity studies (exposure greater than 90 days) in adult mice, rats, rabbits, monkeys and dogs are summarized below. In all the studies considered, aluminum was administered orally in drinking water, in the diet or by gavage. The aluminum salts include lactate, chloride, sulphate, nitrate, hydroxide, citrate, maltolate, fluoride and KASAL (basic sodium aluminum phosphate).

As in the case of the short-term studies, many of the subchronic and chronic toxicity studies did not quantify the concentration of aluminum in the base diet. In those studies where the administered dose was substantially less than the possible baseline dose in the diet, the uncertainty associated with the actual combined dose was increased (see, for example, Krasovskii et al. (1979); Fleming and Joshi (1987); Bilkei-Gorzo (1993); Varner et al. (1993); Varner et al. (1994); Varner et al. (1998); Sahin et al. (1995); Somova et al. (1997); Jia et al. (2001a); Pratico et al. (2002); Abd-Elghaffar et al. (2005); Hu et al. (2005); Becaria et al. (2006); and Li et al. (2006)).

Neurobehavioural effects in adult mice and rats, following oral exposure for 90 days or more, included decreased spontaneous motor activity (Commissaris et al. 1982; Lal et al. 1993; Jia et al. 2001a; Jia et al. 2001b; Hu et al. 2005). The lowest administered dose associated with this effect was 1 mg Al/kg bw/d as observed by Huh et al. (2005) in male Sprague-Dawley rats who received aluminum maltolate at this dose in drinking water over a period of one yearFootnote 15. In contrast Domingo et al. (1996) and Colomina et al. (2002) found no differences in field activity of Sprague-Dawley rats, where animals received an administered dose of 100 mg Al/kg bw/d of aluminum nitrate (with citrate) in drinking water for periods of four to six months. Decreased motor coordination as measured by performance in the rotarod test (Sahin et al. 1995), decreased grip strength, and effects on temperature sensitivity and negative geotaxis (Golub et al. 1992a) were also observed.

Other observed neurobehavioural effects included learning and memory deficits (maze performance, passive avoidance tests) reported by Bilkei-Gorzo (1993), Lal et al. (1993), Gong et al. (2005), Gong et al. (2006) and Li et al. (2006). The lowest administered dose associated with such effects was 6 mg Al/kg bw/d, observed by Bilkei-Gorzo (1993) in Long Evans rats exposed for 90 days to aluminum chloride (plus citrate) via gavage, although there was some ambiguity in the reporting of doses in this study. In contrast, no effects on similar learning or memory tests were observed by Varner et al. (1994), Domingo et al. (1996), Colomina et al. (2002) and von Linstow Roloff et al. (2002). In the study of von Linstow Rolloff et al. (2002) an administered dose of 140 mg Al/kg bw/d was administered to male Lister hooded rats as aluminum sulphate in drinking water.

With respect to body weight, Pettersen et al. (1990), Gupta and Shukla (1995), Colomina et al. (2002) and Kaneko et al. (2004) observed reductions in body weight in aluminum-exposed animals (rodents and dogs) at doses ranging from 25 mg Al/kg bw/d of aluminum maltolate administered in drinking water to mice for up to 120 days (Kaneko et al. 2004) to 94 mg Al/kg bw/d of aluminum nitrate administered in drinking water to rats for 114 days (Colomina et al. 2002). In the Kaneko et al (2004) study, aluminum chloride was administered to another exposure group at the same dose as aluminum maltolate, and no difference in body weight between aluminum-exposed animals and controls was observed. The authors attributed the contrasting observations to the greater bioavailability of aluminum maltolate as compared to chloride, documented as well by the greater accumulation of aluminum in the brain, liver, kidney and spleen in mice exposed to aluminum maltolate.

Histopathological effects reported in rats and mice included increased damaged or abnormal neurons in specific brain regions (for example, cerebral cortex and hippocampus) (Varner et al. 1993; Varner et al. 1998; Abd-Elghaffar et al. 2005), neurofibrillary degeneration and vacuolization of nuclei (Somova et al. 1997), and vacuolated astrocytes and vacuolization of neuronal cytoplasm (Florence et al. 1994). The lowest administered dose in which these effects were observed was less than 1 mg Al/kg bw/d in the Varner et al. (1998) and Varner et al. (1993) studies in which aluminum nitrate and sodium fluoride (to form aluminum fluoride) was administered in drinking water to male Long Evans rats for periods of 45 to 52 weeksFootnote 16.

Petterson et al. (1990) observed mild to moderate histopathological effects in testes, liver and kidney, including hepatocyte vacuolization, seminiferous tubule germinal epithelial cell degeneration and tubular-glomerularnephritis in beagle dogs receiving a dose of 75 mg Al/kg bw/d of sodium aluminum phosphate. In this same study, no significant differences between exposure groups and controls were observed at the lower doses of 4 to 27 mg Al/kg bw/d.

The biochemical endpoints examined in subchronic and chronic experimental studies are considerably varied, as are the methodologies used to investigate these endpoints. The observed effects included a decrease in nitrergic neurons in the somatosensory cortex (Rodella et al. 2006), perturbations in ATPase activity in the brain (Lal et al. 1993; Sarin et al. 1997; Swegert et al. 1999; Silva and Goncalves 2003; Kohila et al. 2004; Silva et al. 2005), induced apoptosis in the brain (Huh et al. 2005), effects on cholinergic enzyme activities (Bilkei-Gorzo 1993; Zheng and Liang 1998; Dave et al. 2002; Zatta et al. 2002; Kohila et al. 2004), increased cytokine levels (Becaria et al. 2006), increased catalytic efficiency of monoamine oxidases A and B (Huh et al. 2005), increased caspase 3 and 12 (Gong et al. 2005; Huh et al. 2005), increased staining for amyloid precursor protein levels (Gong et al. 2005) and amyloid beta (Ab) levels (Pratico et al. 2002), decrease in long-term potentiation in hippocampal slices (Shi-Lei et al. 2005), and alterations in phospholipid and cholesterol levels in the myelin membrane, synaptosomes or the brain (Sarin et al. 1997; Swegert et al. 1999; Pandya et al. 2001; Silva et al. 2002; Pandya et al. 2004). The lowest administered dose associated with significant effects on biochemical endpoints was 1 mg Al/kg bw/d as administered as aluminum maltolate in drinking water for one year (Huh et al. 2005)Footnote 17.

Other biochemical and biophysical effects observed in the brains of aluminum-exposed rodents included alterations in trace metal (Cu, Zn and Mn) metabolism in the brain (Sanchez et al. 1997; Yang and Wong 2001; Jia et al. 2001a; Fattoretti et al. 2003; Fattoretti et al. 2004), altered synapses in the hippocampus and frontal cortex (Jing et al. 2004), increase in area occupied by mossy fibres in the hippocampal CA3 subfield (Fattoretti et al. 2003; Fattoretti et al. 2004), increase (Flora et al. 2003) and decrease (Jia et al. 2001a) in glutathione peroxidase activity, and increase in catalase activity (Flora et al. 2003). Increased lipid peroxidation was reported by Lal et al. (1993), Gupta and Shukla (1995), Sarin et al. (1997), Pratico et al. (2002), Flora et al. (2003) and Kaneko et al. (2004). Jia (2001a), Gupta and Shukla (1995) and Abd-Elghaffar (2005) reported decreased levels of superoxide dismutase, and Jia et al. (2001a) observed increased levels in malondialdehyde. Johnson et al. (1992) observed decreased levels of cytoskeletal proteins (microtubule associated protein-2, spectrin) in the hippocampus and brain stem.

Inhalation exposure

The toxicological literature for subchronic and chronic inhalation exposure studies is limited. ATSDR (2006) and Krewski et al. (2007) report on several studies of durations of six months (six hours a day, five days a week). The most sensitive and best documented endpoints concerned the respiratory system. The observed effects are those commonly associated with particle inhalation exposure (> 600µg/m³), including a thickening of the alveolar walls, and an increase in alveolar macrophages, granulomatous lesions and relative lung weight (ATSDR 2006).

2.4.2.4 Reproductive and developmental toxicity
Oral exposure

The results of 49 studies investigating gestational, lactational and/or post-weaning exposure of rats, mice and guinea pigs to aluminum salts through diet, through drinking water or by gavage are summarized below. The aluminum salts administered in these studies included chloride, nitrate, sulphate, lactate and hydroxide. In a few studies citrate or ascorbic acid was added to enhance absorption of aluminum.

As discussed in sections 2.4.2.2 and 2.4.2.3, the lack of information on base diet for some studies is a major source of uncertainty with respect to the potential combined dose, particularly when the administered dose was low in comparison to the possible base diet dose (for example, Clayton et al. 1992; Ravi et al. 2000). There is also uncertainty associated with reported lowest observed effect levels (LOELs) that are of the same magnitude as the reported LD50 for the administered salt (Johnson et al. 1992; Misawa and Shigeta 1993; Poulos et al. 1996; Llansola et al. 1999).

The most commonly observed neurobehavioural effects in developmental studies included decreased grip strength (Golub et al. 1992b; Golub et al. 1995; Colomina et al. 2005), reduced temperature sensitivity (Donald et al. 1989; Golub et al. 1992b), reduced or delayed auditory startle responsiveness (Misawa and Shigeta 1993; Golub et al. 1994), and impaired negative geotaxis response (Bernuzzi et al. 1986; Bernuzzi et al. 1989a; Muller et al. 1990; Golub et al. 1992b). Decreased activity levels (Cherroret et al. 1992; Misawa and Shigeta 1993), locomotor coordination (Golub et al. 1987; Bernuzzi et al. 1989a; Bernuzzi et al. 1989b; Muller et al. 1990; Golub and Germann 2001b) as well as impaired righting reflex (Bernuzzi et al. 1986; Bernuzzi et al. 1989b) were also observed, although not consistently--refer to Thorne et al. (1987), Golub et al. (1992b), and Misawa and Shigeta (1993). The lowest administered dose at which effects on these endpoints were observed was 100 mg Al/kg bw/d, observed in Wistar rats administered aluminum lactate in the maternal diet during gestation (Bernuzzi et al. 1989b) as well as in Swiss Webster mice administered aluminum lactate in the maternal diet during gestation, lactation and then in the diet of offspring throughout the lifespan (Golub et al. 2000).

The observations on the effects on learning and memory of developmental exposure to aluminum salts also varied considerably. For example, in some studies improved performance in the maze tasks was observed (Golub et al. 2000; Golub and Germann 2001a; Colomina et al. 2005) while in others impaired performance (Golub and Germann 2001b; Jing et al. 2004) or no change (Thorne et al. 1987) was found. Golub and Germann (2001b) observed diminished maze learning in Swiss Webster mice pups when dams were exposed to aluminum lactate in the diet at a combined dose of 50 mg Al/kg bw/d, but not at 10 mg Al/kg bw/d, during gestation and lactation, and pups were exposed via diet for two weeks following weaning. In this experiment, animals (controls and aluminum-exposed) were fed a sub-optimal diet, designed to simulate the usual diet of U.S. women with regard to recommended dietary amounts of trace elements.

The observations of Roig et al. (2006) suggested a biphasic effect on learning in rats exposed to aluminum nitrate during gestation, lactation and post-weaning; in a two-dose study, the low-dose group (50 mg Al/kg bw/d of aluminum nitrate plus citrate in drinking water) performed significantly better in the water maze test than the high-dose group (100 mg Al/kg bw/d), but there was no significant difference between the high-dose group and the controls. With respect to passive avoidance tests, the same group of researchers also reported improved performance in aluminum exposed animals at an administered dose of 100 mg Al/kg bw/d (Colomina et al. 2005).

Developmental exposure of mice and rats to aluminum salts also produced some evidence of disturbances in brain biochemistry, such as alterations in brain lipid contents and increased lipid peroxidation (Verstraeten et al. 1998; Verstraeten et al. 2002; Nehru and Anand 2005; Sharma and Mishra 2006) or decreased lipid peroxidation (Golub and Germann 2000), decreased levels in superoxide dismutase (Nehru and Anand 2005), delayed expression of a phosphorylated neurofilament protein (Poulos et al. 1996), differential effects on choline acetyltransferase activity in various brain regions (Clayton et al. 1992; Rajasekaran 2000; Ravi et al. 2000), decreased serotonin and noradrenaline levels in specific brain regions (Ravi et al. 2000), decreased concentrations of manganese in brain (Golub et al. 1992b; Golub et al. 1993), alterations to signal transduction pathways associated with glutamate receptors and decreased expression of proteins of the neuronal glutamate-nitric oxide-cGMP pathway (Llansola et al. 1999; Kim 2003), and alterations in secondary messenger systems (Johnson et al. 1992). With respect to biochemical endpoints, the lowest administered dose at which effects were measured was approximately 20 mg Al/kg bw/d, observed by Kim (2003) in which male and female Fisher rats received this dose of aluminum chloride in drinking water for 12 weeks prior to mating, after which treatment at this dose continued in dams during gestation and lactation.

Chen et al. (2002), Wang et al. (2002a) and Wang et al. (2002b) reported impairment of synaptic plasticity, as measured by field potentials in the dentate gyrus of the hippocampus. Johnson et al. (1992) reported decreased levels of microtubule associated protein-2 in the brains of rat pups exposed eight weeks following weaning, although no changes in other cytoskeletal proteins were observed. A significant decrease in myelin sheath width was observed in mice pups exposed during gestation, lactation and then through the diet following weaning (Golub and Tarara 1999), and in guinea pig pups exposed prenatally from GD30 to birth (Golub et al. 2002). These effects were observed at administered doses above 85 mg Al/kg bw/d as aluminum chloride in drinking water of Wistar rat dams (Wang et al. 2002a; Wang et al. 2002b; Chen et al. 2002) and 100 mg Al/kg bw/d in the diet of Swiss Webster mice dams (Golub and Tarara 1999).

Although the focus of the majority of the investigations of prenatal exposure was neurodevelopmental toxicity, effects on some reproductive endpoints were reported as well. Golub et al. (1987), Bernuzzi et al. (1989b), Gomez et al. (1991), Colomina et al. (1992), Belles et al. (1999), Sharma and Mishra (2006) and Paternain et al. (1988) reported reduced maternal weight gain, although no change in this parameter was observed by Donald et al. (1989), Golub et al. (1993), Golub et al. (1995) and Golub et al. (1996a), nor was it reported in the other studies. In regard to pup body weight, Sharma and Mishra (2006), Wang et al. (2002a), Llansola et al. (1999), Cherroret et al. (1995), Misawa and Shigeta (1993), Gomez et al. (1991), Paternain et al. (1988), Domingo et al. (1987), Thorne et al. (1987), Golub and Germann (2001a), Colomina et al. (1992), and Bernuzzi et al. (1989a), Bernuzzi et al. (1989b) reported decreases in aluminum-exposed groups, while other studies reported no effects (Donald et al. 1989; Clayton et al. 1992; Golub et al. 1992b; Golub et al. 1993; Golub et al. 1995; Golub et al. 1996a; Colomina et al. 1994; Verstraeten et al. 1998). The lowest administered dose at which effects on reproductive parameters, including fetal growth, were observed was 13 mg Al/kg bw/d (Paternain et al. 1988; Domingo et al. 1987a), in which Sprague-Dawley rat dams received this dose via gavage as aluminum nitrate.

Cherroret et al. (1995) reported decreased plasma concentrations of total proteins and albumin and increased plasma a1 globulins, which the authors attributed to an inflammation process in young rats exposed postnatally by gavage at doses of 100 to 200 mg Al/kg bw/d. The same research group also observed effects on duodenal enterocytes, with a decrease in microvilli width and significant variation in K, Ca, S and Fe concentrations (Durand et al. 1993).

Other observed reproductive/developmental effects included a decrease in the number of corpora lutea and number of implantation sites (Sharma and Mishra 2006) as well as skeletal malformations (Paternain et al. 1988; Colomina et al. 1992; Sharma and Mishra 2006). Colomina et al. (2005) reported a delay in sexual maturation in both males and females, although this effect was produced at different dose levels in the two sexes (at 50 mg Al/kg bw/d in females and at 100 mg Al/kg bw/d in males). Misawa and Shigeta (1993) observed delayed pinna detachment and eye opening in female pups.

No significant maternal or developmental toxicity, as measured by fetal weight gain, reproductive parameters or fetal malformations, was observed by McCormack et al. (1979) at a combined dietary dose of aluminum chloride of 50 mg Al/kg bw/d, nor by Gomez et al. (1990) where 265 mg Al/kg bw/d of aluminum hydroxide was administered to dams via gavage during gestation.

Inhalation and dermal exposure

No studies were identified concerning the reproductive effects of inhalation or dermal exposure to aluminum salts.

2.4.2.5 Carcinogenicity

The literature concerning oral exposure bioassays is very limited. An increase in gross tumours was reported in male rats and female mice in a one-dose study but few study details were reported (Schroeder and Mitchener 1975a, 1975b, as reported in ATSDR 2006). Two other studies reported no increased incidence of tumours in rats and mice exposed orally to aluminum compounds (Hackenberg 1972; Oneda et al. 1994).

No increased tumour incidence was observed in rats following inhalation of alumina fibres at concentrations of up to 2.45 mg/m³ (Krewski et al. 2007).

The International Agency for Research on Cancer did not classify specific aluminum compounds for carcinogenicity, but classified the exposure circumstances of aluminum production as carcinogenic to humans (Group 1) (IARC 1987).

2.4.2.6 Genotoxicity

The genotoxicity of various aluminum compounds is described in detail by Krewski et al. (2007) and ATSDR (2006). Briefly, aluminum compounds have produced negative results in most short-term in vitro mutagenic assays, including the Rec-assay using Bacillus subtilis, in Salmonella typhimurium TA92, TA 98, TA102, TA104 and TA1000 strains (with and without S9 metabolic activation), and in Escherichia coli (see Krewski et al. 2007).

In vitro studies of rat ascites hepatoma cells reported that aluminum chloride could serve as a stimulator for the crosslinking of chromosomal proteins (Wedrychowski et al. 1986a, 1986b, as reported in Krewski et al. 2007, ATSDR 2006). Studies on human blood lymphocytes showed that aluminum chloride could induce positive responses for both micronuclei formation and sister chromatid exchange (see Krewski et al. 2007).

More recently Lima et al. (2007) investigated the genotoxic effects of aluminum chloride in cultured human lymphocytes. Comet assay and chromosome aberrations analysis were used to evaluate DNA-damaging and clastogenic effects of aluminum chloride at different phases of the cell cycle. All tested concentrations (5 to 25 µM aluminum chloride) were cytotoxic, reduced the mitotic index, induced DNA damage and were clastogenic in all phases.

Roy et al. (1991) administered doses of aluminum sulphate and potassium aluminum sulphate in drinking water to male rats at doses ranging from 17 to 171 mg Al/kg bw/d for up to 21 days. The frequency of abnormal cells increased in direct proportion to both the dose and the duration of exposure to the aluminum salts. Most aberrations were chromatid breaks, with translocations recorded at higher doses.

In a recent review of the safety of aluminum from dietary intake, EFSA (2008) summarized indirect mechanisms that might explain the genotoxic effects observed in experimental systems. The proposed mechanisms included cross-linking of DNA with chromosomal proteins, interaction with microtubule assembly and mitotic spindle functioning, induction of oxidative damage, and damage of lysosomal membranes with liberation of DNAase to explain the induction of structural chromosomal aberrations, sister chromatid exchanges, chromosome loss and formation of oxidized bases in experimental systems. EFSA (2008) suggested that these indirect mechanisms of genotoxicity, occurring at relatively high levels of exposure, would not likely be of relevance for humans exposed to aluminum via the diet.

2.4.3 Human studies

In this section, information on the potential human health effects associated with aluminum exposure is briefly summarized with the goal of describing the range of potential effects. As such, various exposure routes are considered in order to identify the possible target organs. This information includes data from case studies, epidemiological investigations into the potential health effects of exposure to aluminum in drinking water, occupational investigations of exposure to aluminum dust and welding fumes, and exposure to aluminum via vaccines and of dermal application of aluminum-containing antiperspirants.

In section 3.2.2 an evaluation of these health effects is presented, in order to: (a) identify critical effects; and (b) determine which, if any, of the human studies may be used to estimate the dose-response relationship. The latter determination is based on the strength of the available evidence and the relevance of the studies to environmental exposure in the general Canadian population.

2.4.3.1 Human case studies of exposure to aluminum

Human cases studies of aluminum toxicity have been well documented for specific medical conditions, most frequently in patients with renal impairment undergoing dialysis with aluminum-contaminated dialysate or receiving medications with elevated aluminum concentration. A small number of case studies or investigations have focused on children and pre-term infants receiving parenteral nutrition. Although the effects in particular sub-groups of susceptible individuals are not representative of exposure conditions for the general population, they are presented in order to identify the target organs of aluminum exposure. A more detailed discussion of these human case studies is presented in the comprehensive reviews InVS-Afssa-Afssaps (2003) and Krewski et al. (2007). As well, a case study is described below in which exposure to aluminum was associated with the accidental discharge of aluminum into the municipal water supply.

Aluminum toxicity in patients with renal impairment

Historically, patients undergoing dialysis treatment were exposed to aluminum through the water used to prepare dialysis solutions and from aluminum compounds prescribed as phosphate binders (Krewski et al. 2007). Today, this exposure is strictly controlledFootnote 18. However, in the past, many cases of aluminum-induced encephalopathy, resulting in alterations in behaviour and memory, speech disorders, convulsions and muscle-twitching occurred in dialysis patients (Foley et al. 1981; Alfrey 1993). In cases of intoxication, the aluminum was introduced into the systemic circulation through the dialyzing membrane (in hemodialysis) or abdomen (in peritoneal dialysis) thus bypassing the gastrointestinal barrier, and was therefore completely available at the cellular level. The effects of elevated aluminum exposure in dialysis patients has provided clear evidence for the neurotoxicity of aluminum in humans.

Researchers have also identified cases of individuals with impaired renal function who, because of their reduced capacity to eliminate aluminum and chronic high exposure to aluminum-containing medications, also developed encephalopathy, even though they were not undergoing dialysis (Foley et al. 1981; Sedman et al. 1984; Sherrard et al. 1988; Moreno et al. 1991). A fatal case of aluminum-induced encephalopathy occurred in a patient with chronic renal failure who did not have dialysis treatment, but who consumed large doses of aluminum-containing antacids (Zatta et al. 2004).

Other toxic effects of aluminum observed in dialysis-exposed patients include haematological effects such as anaemia (Bia et al. 1989; Yuan et al. 1989; Shah et al. 1990; Caramelo et al. 1995) and skeletal toxicity (osteomalacia and osteitis fibrosis) (Mathias et al. 1993; Jeffery et al. 1996; Ng et al. 2004).

Aluminum exposure via intravenous nutritional support

Klein (2005) reviewed the human evidence regarding the effects of aluminum exposure via solutions used for intravenous nutritional support with regard to effects on bone (osteomalacia) and the central nervous system. With respect to parenteral nutrition, infants may be a particularly sensitive sub-group because of the immaturity of the blood-brain barrier and renal excretory mechanisms. Bishop et al. (1997) investigated cognitive impairment in pre-term infants in relation to parenteral nutrition. In a randomized trial the researchers found that performance in neurodevelopmental testing conducted at 18 months was significantly better in 92 pre-term infants who had received a low-aluminum nutritional solution as compared to 90 pre-term infants receiving a standard solution with higher aluminum content. No follow-up testing that evaluated cognitive performance in the children of this cohort as they aged was identified.

Investigation of aluminum exposure associated with contamination event in Camelford, UK

Exley and Esiri (2006) reported an unusual case of fatal dementing illness in a 58-year-old woman, resident of Camelford, Cornwall, in the United Kingdom. Fifteen years earlier, at the age of 44 years, this person was exposed to high concentrations of aluminum sulphate in drinking water, which had been accidentally discharged in the drinking water supply of the region. During this event, up to 20,000 people were exposed to aluminum concentrations in drinking water varying from 100 to 600 mg/L. At the autopsy of the woman, a rare form of sporadic early-onset b-amyloid angiopathy in the cerebral cortical and leptomeningeal vessels, and in leptomeningeal vessels over the cerebellum was identified. Coincident high concentrations of aluminum were also found in the severely affected regions of the cortex. To date, this remains the only documented case. Exley and Esiri (2006), who reported this case, state that the role of aluminum is uncertain but may be clarified through future research in similarly exposed and unexposed populations (controls).

2.4.3.2 Epidemiological studies of aluminum exposure via drinking water

By the end of the 1980s, four epidemiological studies with an ecological design (that is, using group rates of exposure and disease) had reported positive associations between the concentration of aluminum in drinking water and the occurrence of Alzheimer's disease (AD) or of dementia (Vogt 1986; Martyn et al. 1989; Flaten 1990; Frecker 1991). These observations resulted in further research into the relationship of aluminum in drinking water and various dementia syndromes, particularly AD.

Epidemiological studies based on observations of individuals were conducted in the 1990s with the aim of investigating the association between AD or other cognitive dysfunctions and exposure to aluminum in drinking water. Health Canada published a comprehensive review of epidemiological studies in Guidelines for Canadian Drinking Water Quality - Technical Documents: Aluminum (Health Canada 1998b) and in the State of Science Report of 2000. The discussion presented below summarizes the information presented in the previous reviews and presents more recently published findings of the eight-year follow-up analysis of a large cohort in southwestern France (Rondeau et al. 2000; Rondeau et al. 2001). The study designs and findings of the relevant epidemiological studies are presented in Table B1 (Appendix B). These data have also been described in detail in Krewski et al. (2007) and InVS-Afssa-Afssaps (2003). Analysis of the epidemiological database and its applicability in a quantitative risk assessment is presented in the Hazard Characterization of this assessment (section 3.2.2.1).

Twelve studies are presented in Table B1, based on case-control, cross-sectional, or longitudinal designs. The observations from two Ontario case-control studies are drawn from the same study population--the Ontario Longitudinal study of Aging (LSA)--and all the French studies were based on observations from the "Principal lifetime occupation and cognitive impairment in a French elderly cohort" or PAQUID cohort. However, the LSA and PAQUID study populations differ with respect to the case definition and the manner of diagnosis of disease. In the PAQUID investigations, the earlier studies used a case-control design whereas the more recent studies by Rondeau et al. (2000) and Rondeau et al. (2001) used a cohort incidence analysis.

Positive findings for an association between aluminum exposure and AD or other neurological dysfunctions were found to be statistically significant (p < 0.05) in seven of the twelve studies, although the strength and significance of these associations depended on how the data were analysed (see Appendix B) These seven studies were carried out in Ontario (Neri and Hewitt 1991; Forbes et al. 1992; Forbes et al. 1994; Forbes et al. 1995a; Forbes et al. 1995b; Neri et al. 1992; Forbes and Agwani 1994; Forbes and McLachlan 1996; McLachlan et al. 1996), in Quebec (Gauthier et al. 2000) and in France (Michel et al. 1991; Jacqmin et al. 1994; Jacqmin-Gadda et al. 1996; Rondeau et al. 2000; Rondeau et al. 2001).

In Ontario, a series of analyses was conducted on the LSA cohort to investigate the relationship between the concentration of aluminum in drinking water and cognitive impairment, as established by interviews and questionnaires (Forbes et al. 1992; Forbes et al. 1994; Forbes et al. 1995a; Forbes and Agwani 1994). These authors observed statistically significant associations only when they controlled their analyses according to certain physical-chemical parameters of water, such as fluoride, pH, and silica. Since the methods of interviews and questionnaires for characterizing cognitive functions were deemed to be insufficiently specific for accurately detecting neurological impairments, Forbes et al. (1995b) and Forbes and McLachlan (1996) consulted death certificates from individuals on the LSA cohort and examined the association between aluminum in drinking water and AD or presenile dementia as categorized by the corresponding ICDFootnote 19 codes. Positive relationships between aluminum and AD and presenile dementia were reported with and without adjustments with different water quality parameters. For instance, some of the highest risks for AD were observed when high concentrations of aluminum (≥ 336 µg Al/L) were combined with high pH (≥ 7.95), low levels of fluoride (< 300 µg/L) or low levels of silica (< 1.5 mg/L).

Neri and Hewitt (1991) and Neri et al. (1992) reported a significant dose-response relationship between AD or presenile dementia and aluminum using hospital discharge records from Ontario, and by matching cases and controls according to age and sex. Another study from Ontario was a case-control analysis from the Canadian Brain Tissue Bank cohort in which AD was confirmed by histopathological criteria (McLachlan et al. 1996).

Although all studies from Ontario assessed the exposure of aluminum based on the data of the water quality surveillance program of the Ontario Ministry of the Environment, only McLachlan et al. (1996) evaluated the past exposure to aluminumFootnote 20. However, in the McLachlan et al. (1996) study, the analysis was not controlled for potential confounders and modifying factors (for example, age, sex, education and occupation), and the significant positive associations were not adjusted for other chemical or physical parameters in water.

The single study from Quebec was a case-control analysis of AD and exposure to various aluminum species in residential drinking water (Gauthier et al. 2000). The diagnosis of AD was based on a three-step procedure to discriminate between AD and other neurological disorders. In addition to controlling for a number of confounding factors as well as the aluminum speciation, these authors took into account historical exposure to aluminum in drinking water. Gauthier et al. (2000) reported 16 odds ratios (OR) but observed only one significant positive association (that is, OR > 1), which was related to the concentration of monomeric organic aluminum in drinking water. This significant association was found, however, when only current exposure was considered, and not for long-term exposure, which would be expected to be more biologically-relevant.

The three studies conducted on populations from the United Kingdom showed no significant association between aluminum concentration in drinking water and neurological dysfunction, following adjustment for sex and age (Wood et al. 1988; Forster et al. 1995; Martyn et al. 1997), but none of these authors adjusted their statistical tests according to the physical-chemical properties of the drinking water. The health outcome in the two case-control studies was AD, diagnosed by a three-step procedure for including cases of presenile dementia (Forster et al. 1995) or by a clinical diagnosis using unspecified criteria (Martyn et al. 1997). This latter study, which took into account past exposure, also did not observe differences between cases and controls when the analyses were restricted to subjects exposed to low levels of silica in drinking water (< 6 mg/L). The cross-sectional study of Wood et al. (1988) was based on data collected from patients from northern England with hip fractures, for whom dementia was evaluated (no information about the diagnostic tests).

The study from Switzerland (Wettstein et al. 1991), which was a cross-sectional examination of mnestic skills in octogenarians from Zurich and aluminum in drinking water, also reported no significant associations when controlling for socio-economic status, age, and education. It should be noted that the high-exposure district in this study had drinking water with a mean aluminum concentration of 98 µg/L. Thus the analysis was carried out for a drinking water supply that was generally lower in aluminum than the drinking water supplies considered in the other investigations.

All the studies from France were based on the PAQUID cohort. The studies of Michel et al. (1991) and Rondeau et al. (2001), reported significant positive associations between the exposure to aluminum in drinking water and the occurrence of AD or dementia diagnosed by a two-step procedure, whereas the positive associations reported by Jacqmin et al. (1994) and by Jacqmin-Gadda et al. (1996) were based on the scores of the Mini-Mental State Examination (MMSE). The results of Michel et al. (1991) have been discounted, however, because of a reliance on potentially unreliable historical information on drinking water concentrations (Jacqmin et al. 1994; Smith 1995; WHO 1997).

Jacqmin et al. (1994) and Jacqmin-Gadda et al. (1996) analysed the same database collected from the PAQUID cohort in different ways, with inconclusive results. The first study included an initial report of the effect of pH on the association between aluminum and cognitive impairment (Jacqmin et al. 1994). Without considering the effect of the pH-aluminum interaction, these authors reported a positive association between aluminum and cognitive impairment, whereas consideration of this interaction resulted in a negative association. These results remained statistically significant only if occupation was included in the logistic regressions. Jacqmin-Gadda et al. (1996) expanded their analyses to include the levels of silica in drinking water. While their results indicate a protective effect of aluminum against cognitive impairment with high level of silica (≥ 10.4 mg/L) and high pH (≥ 7.5), the consideration of the interaction of aluminum and silica in their logistic regression suggests an adverse effect of aluminum on neurological functions.

Rondeau et al. (2000) retained the unimpaired subjects in the studies of Jacqmin et al. (1994) and Jacqmin-Gadda et al. (1996), and evaluated the incidence of dementia and AD one, three, five and eight years after the initial MMSE. This follow-up analysis reported a positive association between aluminum and AD or dementia, after adjustment for age, sex, education and place of residence as well as for consumption of wine and bottled mineral water. This study addressed some of the limitations of previous epidemiological investigations by adjusting for the potential confounders, and while exposure levels were not weighted according to residential history, residential history was considered. At baseline, 91% of the subjects had lived more than ten years in the same parish, with a mean length of residence of 41 years. A total of 3,401 participants were included in the study at baseline, although only 2.6% of the subjects were exposed to an aluminum concentration greater than 100 µg/L. Nonetheless, the associations between aluminum in drinking water and dementia, and aluminum in drinking water and AD, were highly significant. Only two exposure groups (< 100 µg/L or > 100 µg/L) were defined in the principal analysis and no dose-response relationship was found when exposure categories were more finely divided.

Many of the epidemiological studies investigating the association between aluminum in drinking water and the development of cognitive impairment or AD did not control for important potential confounders or modifying factors, or did not adequately characterize past exposure. The Rondeau (2000) study addressed some of these limitations. However, the subjects in the cohort were not generally exposed to high levels of aluminum (97% of subjects exposed to less than 100 µg/L), and within the limited exposure range, no dose-response relationship was observed.

2.4.3.3 Epidemiological investigations of exposure to aluminum in antacids, antiperspirants or food

Only very weak or no associations have been found between repeated exposures to aluminum in antacids and AD in a number of analytical epidemiological studies (Heyman et al. 1984; Graves et al. 1990; Flaten et al. 1991; CSHA 1994; Forster et al. 1995; Lindsay et al. 2002) Positive associations between AD and the use of aluminum containing antiperspirants were reported in two case-control studies, but the interpretation of the results is difficult due to methodological limitations of the studies (for example, missing data, and misclassification due to varying brands and subtypes of antiperspirant with varying aluminum contents) (Graves et al. 1990; CSHA 1994). This positive observation, however, was not supported by a follow-up study on the CSHAFootnote 21 cohort (Lindsay et al. 2002); the results show that regular use of antiperspirant did not increase the risk of AD.

Rogers and Simon (1999) conducted a pilot study to examine dietary differences in individuals with AD and matched controls (n = 46: 23 subjects, 23 controls). The exposure assessment was based on questionnaires to determine past dietary habits. According to the authors, there may be an association between AD and the consumption of foods containing high levels of aluminum food additives. However, the sample size was very small and the association was statistically significant only for one category of food (pancake, waffle and biscuit).

2.4.3.4 Epidemiological investigations of exposure to aluminum in vaccines

Aluminum adjuvants are included in some vaccines to enhance and extend the immune response of some antigens. Aluminum hydroxide and phosphate salts as well as aluminum sulphate can be used as an adjuvant (Eickhoff and Myers 2002).

Possible associations between AD and historical exposure to vaccines have been investigated in the CSHA cohort (Verreault et al. 2001). Exposure to conventional vaccines appears to lower the risk of developing AD. After adjustments for age, sex and education, the ORs were 0.41 (95% CI 0.27-0.62) for the diphtheria or tetanus vaccines, 0.60 (95% CI 0.37-0.99) for the poliomyelitis vaccines and 0.75 (95% CI 0.54-1.04) for the influenza vaccine. Except for the influenza vaccine, all others contain aluminum-adjuvants (Eickhoff and Myers 2002).

The possible links between the hepatitis B vaccine, which contains aluminum-adjuvants, and the risk of demyelinating diseases such as multiple sclerosis (MS) have been investigated in France (Touze et al. 2000; Touze et al. 2002), England (Sturkenboom et al. 2000; Hernan et al. 2004), the U.S. (Zipp et al. 1999; Ascherio et al. 2001), Canada (Sadovnick and Scheifele 2000) and Europe (Confavreux et al. 2001). Only the study of Hernan et al. (2004) observed a significant positive association between MS and the hepatitis B vaccine, but no association between MS and the tetanus or influenza vaccines, which also contain aluminum adjuvants.

2.4.3.5 Epidemiological investigations of occupational exposure to aluminum

Subclinical neurological effects have been observed in a number of studies of workers chronically exposed to aluminum (aluminum potroom and foundry workers, welders, and miners). Many of these studies involved small numbers of workers and involved the assessment of exposure based on occupation rather than measured airborne aluminum concentrations, and most involved mixed exposures to various dusts and chemicals. Endpoints examined in different studies varied and for those that were similar, results were not always consistent. The types of neurological effects observed included impaired motor function (Hosovski et al 1990; Sjogren et al 1996; Kilburn 1998), decreased performance on cognitive tests (attention, memory, visuospatial function) (Hosovski et al 1990; Rifat et al 1990; Bast-Pettersen et al 1994; Kilburn 1998; Akila et al 1999), subjective neuropsychiatric symptoms (Sjogren et al 1990; White et al 1992; Sim et al 1997) and quantitative electroencephalographic changes (Hanninen et al 1994).

In one case-control study from England (Salib and Hillier 1996) and two from the U.S. (Gun et al 1997; Graves et al 1998), the relationship between the occurrence of AD and occupational exposure to aluminum was investigated. In each study, disease status was defined by standard criteria (for example, NINCDS-ADRDA and/or DSM)Footnote 22, and exposure to airborne aluminum (for example, welding fumes, dusts and flakes) was assessed through occupational history questionnaires administered to informants. In none of these studies was there a significant association between occupational exposure to airborne aluminum and AD.

A four-year longitudinal study investigated neurobehavioural performance in 47 aluminum welders in the train and truck construction industry, with a control group drawn from assembly workers in the same industry (Kieswetter et al 2007). Exposure to aluminum in dust was assessed through total dust collected on filter samples attached to the welders' helmets as well as through biomonitoring (aluminum in plasma and urine) at the time of neurobehavioural testing (start of investigation, after two years, after four years). The battery of neurobehavioural tests included an evaluation of cognitive abilities, psychomotor performance, attention and memory. This study used a small number of participants to explore the potential use of different biomonitoring measures, dust levels and exposure duration to predict performance in neurobehavioural tests. The study was not designed to find a relationship if one existed, but rather to explore the use of different exposure measures. Although exposure to aluminum among the welders was considered to be high in comparison to other occupational studies of aluminum (88 to 140 µg Al/g creatinine in urine, or approximately 103 to 164 mg Al/L)Footnote 23, no association between exposure and neurobehavioural performances was found.

A meta-analysis was conducted for nine investigations of occupational aluminum exposure and neurobehavioural performance, with a total of 449 exposed subjects with mean urinary aluminum concentrations of 13 to 133 µg Al/L (Meyer-Baron et al 2007). Even if almost all effect sizes indicated an inferior neurobehavioural performance of the exposed group to aluminum, only one out of ten performance variables (the digit symbol test) was statistically significant. However, the statistical significance of the digit symbol results relationship to aluminum exposure was reduced when one study, in which the biomonitoring measure was estimated on the basis of an uncertain conversion factor, was excluded from the analysis. The authors concluded that with respect to occupational exposure, as indicated by urinary concentrations of less than 135 µg Al/L, there is concurring evidence of an impact on cognitive performance and acknowledge that international standardization for exposure is needed.

2.4.4 Mode of action of toxic effects of aluminum

Information related to possible modes of action by which aluminum affects the nervous system, as explored in animal and human studies, has been discussed in a number of recent reviews (Strong et al 1996; Savory 2000; Kawahara 2005; ATSDR 2006; Savory et al 2006; Krewski et al 2007; Shcherbatykh and Carpenter 2007; Goncalves and Silva 2007). In addition, Jeffery et al (1996) and Krewski et al (2007) consider the mode of action in relation to bone and hematopoietic tissue.

The mechanism of aluminum neurotoxicity is an area of active research, with multiple lines of investigation. The purpose of the present discussion is to briefly summarize the areas of investigation relating to mode of action of aluminum toxicity, as mostly tested in laboratory rodents or in vitro studies, and present the range of views regarding the relevance of these data to human neurodegeneration, and particularly the development of AD.

Neurotoxic effects

There is evidence from studies in both laboratory animals and humans that absorbed aluminum is distributed to the brain, particularly the cerebral cortex and hippocampus. For example, the accumulation of aluminum in the brains of adult mice, rats and monkeys from the exposed groups was reported in 23 studies of neurological effects of orally administered aluminum (described in section 2.4.2)Footnote 24. Increased aluminum in the brains of pups exposed only during pregnancy was observed by Sharma and Mishra (2006), but not by others (Colomina et al 2005; Golub et al 1992b). Other studies of prenatal exposure in which exposure continued through lactation also reported increased aluminum in the brain (Wang et al 2002a; Chen et al 2002; Golub et al 1993). In contrast, Golub et al (2000) observed decreased aluminum levels in the brains of mice exposed during gestation, lactation and through their lifespan.

Other research documenting the distribution of aluminum in the brain is described in section 2.3.3.2.

The research on aluminum neurotoxicity in laboratory animals has generally focused on the following interrelated categories of biochemical and cellular effects:

  • peroxidation of membrane lipids and other sources of oxidative stress;
  • increased inflammatory response;
  • alterations in the lipid/phospholipid composition of myelin, with consequent effects on neurotransmission and synaptic function;
  • impaired glucose metabolism;
  • effects on neurotransmission, including cholinergic and glutamatergic systems;
  • alterations to second messenger systems (for example, inositol triphosphate, cAMP and Ca2+);
  • accumulation of intracellular calcium;
  • accumulation of mitochondrial Ca2+, resulting in release of cytochrome c and subsequent apoptosis;
  • perturbation in the distribution and homeostasis of essential metals with potential adverse metabolic effects;
  • alteration of phosphorylation level of neurofilaments, including phosphorylation of tau-protein, and resulting neurofibrillary tangle formation;
  • inhibition of axonal transport;
  • accumulation of amyloidβ peptide;
  • alterations in gene expression and binding to DNA;
  • alterations to the permeability of the blood-brain barrier.

There has been some effort to integrate the evidence for the above biochemical effects into a common mechanism, or at least a group of mechanisms of action for the neurotoxicity of aluminum (for example, see Kawahara (2005) and Shcherbatykh and Carpenter (2007)). Strong et al (1996) argued that "a single unifying mechanism of aluminum neurotoxicity that will encompass all the potential means by which aluminum acts at the cellular level probably does not exist." These authors did, however, propose the following general categories by which aluminum neurotoxicity may be characterized, as a means for focusing future research on mechanisms of action:

  • the induction of cytoskeletal pathology in the form of neurofilamentous aggregates: the mechanisms of this induction include those at the level of gene expression to altered post-translational processing (phosphorylation or proteolysis) of neurofilaments;
  • alterations in cognition and behaviour in the absence of cytoskeletal pathology but with significant neurochemical and neurophysiological modifications: these include effects on cholinergic activity, signal transduction pathways and glucose metabolism;
  • developmental neurotoxicity: research into this lifestage could focus on whether the mechanisms of action of aluminum that have led to neurobehavioural alterations in the developing fetus are similar to those responsible for toxicity in the adult as well as the nature of these alterations (permanent versus transient).

The relationship between the mechanism of aluminum neurotoxicity in animals and to the potential mechanism in AD remains an important topic of discussion. This is a complex debate as the basic cellular mechanism for AD is not clear. The presence of senile plaques composed of Ab peptides in the brains of individuals with AD is well-documented, but the means by which these peptides produce neurotoxicity is not known (Marchesi 2005). Superimposed on the debate on the mechanisms for AD is the controversy as to whether environmental exposure to aluminum could contribute to the development of AD. The recent literature includes arguments across the spectrum, from the view that no compelling evidence for the “aluminum hypothesis” exists today (Becking and Priest 1997; Wisniewski and Lidsky 1997) to the view that the different animal and epidemiological evidence suggest that environmental aluminum may indeed be an important contributing factor for AD and that it is important not to prematurely reject this hypothesis (Yokel 2000; Gupta et al 2005; Kawahara 2005; Exley 2006; Miu and Benga 2006; Savory et al 2006). The proponents of further investigation into the role of aluminum in the development of AD cite the following lines of evidence, in addition to the epidemiological evidence described in section 2.4.3.2. For each of these points, the counter arguments that have been put forward are also noted:

  • Increased aluminum in the whole brains of AD individuals at autopsy, as compared to age-matched non-AD brains, has been observed in some studies, although not in others. Investigations focusing on the measurement of aluminum in the senile plaques and neurofibrillary tangles of AD brains have also produced variable results, possibly as a result of difficulties and differences of the analytical methods used (Environment Canada and Health Canada 2000; Yokel 2000).
  • Aluminum injected into the brain or spinal cord of certain species (for example, rabbit, cat, guinea pig and ferret) produce effects that have some similarities to AD pathology, although there are significant differences as well.

    For example, abnormally phosphorylated tau is the principal protein of the paired helical filaments that make up the neurofibrillary tangles that are characteristic of AD. Aluminum-induced phosphorylation of tau protein has been demonstrated in somein vitro and in vivo studies (Yokel 2000; Savory et al 2006). Yet, although aluminum induces neurofilament aggregates in model species, these differ structurally from neurofibrillary tangles that are diagnostic of AD in humans.

    The deposition of senile plaques, a hallmark of AD, is not observed in animal models, but increased immunoreactivity to Aβ and its parent molecule, amyloid precursor protein, via aluminum has been demonstrated in both in vitro and in vivo studies (Environment Canada and Health Canada 2000).

  • Dialysis encephalopathy (see section 2.4.3.1 for discussion) is clearly recognized as resulting from aluminum intoxication. This condition has provided clear evidence for the neurotoxicity of aluminum in humans. Nonetheless, the very different clinical symptoms and progression of the two diseases as well as their differing pathologies have been cited as evidence of a lack of causal relationship between aluminum and AD (Wisniewski and Lidsky 1997).

    In a study investigating the morphological alterations in the brains of 50 patients with a history of chronic renal failure and long-term hemodialysis, Reusche et al (2001) evaluated the degree of changes generally associated with dialysis-associated encephalopathy (DAE) (Al-containing inclusions in glia and neurons) and those associated with AD (βA4 amyloid and AD-type neurofibrillary tangles), and compared to the brains of age-matched patients with a record of heart attacks or carcinomas, and without evdidence of neurological disease.  Cumulative aluminum ingestion was estimated through records of prescription of aluminum-containing drugs (dialysis was not considered a significant source of aluminum exposure in these patients). A statistically significantly association between the degree of DAE-related morphological change and cumulative aluminum ingestion was observed, whereas no such association was observed with respect to the less frequently observed AD-type morphological change, and no significant difference between controls and patients was observed with respect to AD morphology. The authors concluded that exposure to aluminum through long-term ingestion of aluminum-containing drugs does not result in AD morphological changes
Bone toxicity

In the case of osteomalacia associated with aluminum exposure, two distinct mechanisms of actions are recognized (ATSDR 2006). Firstly, the oral exposure to high levels of aluminum can produce a complex with dietary phosphorus, impairing gastrointestinal absorption of this element necessary for bone mineralization. Secondly, the osteomalacia associated with increased bone concentrations of aluminum, principally located at the mineralization front, is associated with increased mineralization lag time, increased osteoid surface area, low parathyroid hormone levels, and elevated serum calcium levels (ATSDR 2006).

Hematopoetic tissue

Among patients with chronic renal failure who receive dialysis treatment, some individuals will develop a hypochromic microcytic anemia, the severity of which correlates with the plasma and red blood cell aluminum levels and can be reversed by terminating exposure to aluminum or by aluminum chelation with desferrioxamine (Jeffery et al. 1996). While the mechanism for this effect in dialysis patients is not known, Jeffery et al. (1996) suggest that it may be aluminum interference with iron metabolism, possibly through disruption in cellular transfer of iron to ferritin to heme.

Report a problem or mistake on this page
Please select all that apply:

Thank you for your help!

You will not receive a reply. For enquiries, contact us.

Date modified: