Lead toxicosis is a serious and potentially fatal disease in companion animals.1–3 Clinical signs are variable, but neurologic and gastrointestinal disturbances alone or in combination are frequently reported and can include seizures, behavioral changes, vomiting, anorexia, and diarrhea.2,4–7 Although lead can be found naturally in the environment, most cases of lead poisoning are the result of anthropogenic environmental contamination.1,3–5,8 Lead-containing plumbing materials, ammunition, fishing weights, toys, and storage batteries as well as the manufacturing and recycling factories for these products are well-known sources of exposure.3–5,8 Furthermore, prior to 1955, household paint contained up to 50% lead, and prior to 1975, lead was a common gasoline additive.9,10 As a result of these industrial practices, lead intoxication was once the most commonly reported accidental poisoning in companion animals.11 In fact, much of the current understanding of the epidemiology, clinicopathologic features, and treatment of lead poisoning in companion animals originated from reports in this time period from the late 1960s through the early 1990s.1–8
Federal regulations have been implemented to reduce environmental lead contamination. Allowable lead concentrations in household paint were reduced to 1% in 1971, and lead-containing household paint was banned in 1978.9 Similarly, allowable concentrations of lead additives in gasoline were substantially reduced in 1985, and leaded gasoline was banned by 1996.10 Because of these actions and others, the prevalence of lead poisoning in the United States has dramatically decreased.12 Although lead poisoning in dogs could be underreported and the overall prevalence is unknown, published reports13,14 of lead toxicosis in dogs in the United States are limited to case studies in the past 20 years.
In the city of Flint, Mich, lead contamination of the water supply resulted in state and federal state-of-emergency declarations in January 2016.15,16 The city of Flint had used water supplied by Lake Huron for decades, but in April 2014, the water supply was switched to the Flint River as a cost-saving measure pending construction of a new water pipeline to Lake Huron. Water from the river was more corrosive to existing plumbing, compared with water supplied by the lake, and corrosion inhibitors were not added to the water supply at that time.15,17 This, coupled with the city's aging infrastructure and extensive use of lead pipes and solder, resulted in contamination of the city water supply. For > 1 year, many residents were exposed to lead-contaminated water, and the impact was greatest in socioeconomically disadvantaged neighborhoods.15 Given the detrimental effects of lead on behavior and cognition, the long-term medical and societal impacts of this crisis have yet to be realized, and it is estimated that recovery costs could exceed $1 billion.15,18
Environmental toxicants such as lead often affect multiple species; companion animals can serve as sentinels for human exposure, and the reverse is also true.19,20 Although most relief efforts in the city of Flint focused on human health, pets also were at risk of exposure. This exemplifies the importance of a one-health outlook and a global approach to human and animal well-being in a public health crisis.21 The purpose of the study reported here was to investigate BLCs in dogs residing in Flint (including immediately adjacent areas outside of city wards but receiving the same water supply) via free screening clinics held between February and May 2016. In addition, we sought to investigate potential associations of BLCs with age, water source, and clinical signs in these dogs. The primary study hypothesis was that BLCs would be greater in dogs residing in Flint than in a control population of dogs.
Materials and Methods
Dogs and study design
Dogs residing in Flint, Mich, and immediately adjacent areas (test population) and dogs residing in East Lansing, Mich, and immediately adjacent areas (control population) were recruited for participation in the cross-sectional design study. The East Lansing area was targeted for the designated control population because the city and its surrounding areas were supplied by well water that consistently tested below reportable limits for lead concentrations.22 The project was approved by the Michigan State University Institutional Animal Care and Use Committee.
Test population—Free lead-screening clinics for dogs living in Flint, Mich, were held at several locations between February 20 and May 14, 2016. State and federal state-of-emergency declarations were still in effect at this time; a timeline of events associated with the Flint water crisis is available from another source.23 Testing clinics were established through collaboration with local community partners to ensure trust between dog owners and study investigators, given the sensitive political environment.24 In total, 6 clinics were held at 5 different locations in the west, north, central, and southeast parts of the city (the county animal control facility, 3 local churches, and the county Humane Society [which hosted 2 clinics]). The sole inclusion criterion was residence within the greater city of Flint regions that received their water supply from the Flint River. This included the 9 city wards of Flint as well as regions that were not contained within city wards but used the city water supply. Dogs residing outside these areas were excluded from analysis. The lead-screening clinics were promoted through various means, including social media, newspapers, television and radio announcements and interviews, and flyer distribution through schools, churches, and other outlets.
Informed consent was obtained from all owners of test population dogs prior to study enrollment. Dog owners were required to complete a questionnaire (Supplementary Appendix S1, available at avmajournals.avma.org/doi/suppl/10.2460/javma.251.8.912) at the time of blood sample collection to provide demographic and clinical information for each dog enrolled. The questionnaire was available in English and Spanish languages, and volunteers were present at each clinic to assist dog owners with any difficulties related to reading the questionnaire. Demographic information obtained with the questionnaire included the dog's age, breed, reproductive status, address of residence, length of time at that address (≤ 12 vs > 12 months), and previous address. The questionnaire solicited information on drinking water sources used for the dog (tap water, bottled water, or filtered water; whether an outdoor spigot was used; and whether other water sources were accessible) as well as the length of time that those water sources were used. Finally, owners were asked to provide subjective assessments of changes in their dog's behavior, to indicate any other perceived abnormalities, and to characterize these if applicable. All screening clinics were staffed by 8 to 16 members of the Michigan State University College of Veterinary Medicine, who assisted with enrollment, animal handling, sample collection, and sample processing.
Control population—Dogs undergoing routine venipuncture for any reason (eg, as part of a wellness examination, for preanesthetic assessment, for genetic screening, or for disease screening or monitoring) at the Veterinary Medical Center of the Michigan State University College of Veterinary Medicine during the same time period as the test population were eligible for inclusion in the control population, provided that they resided in East Lansing, Mich, and immediately adjacent areas. Screening of the dog's blood for lead and other elements was performed at no additional cost to the owners. Medical records were reviewed to retrieve demographic and clinical data for comparisons with the test population. Owners of most dogs (those evaluated by the primary care service) completed the same questionnaire used to gather data for test dogs.
Sample collection and analysis
Blood samples (3 to 5 mL) were collected from all dogs via venipuncture. Samples were placed in 2 separate EDTA-containing blood collection tubes, with 1 used for evaluation of hematologic variables and 1 used for BLC measurement. Blood smears were made, and RBC morphology was assessed at the Michigan State University clinical pathology laboratory. Other hematologic data were obtained with an automated analyzer.a
Blood concentrations of multiple elements were measured; only lead concentration data were used in the study. The BLCs were measured in whole blood at the Michigan State University Diagnostic Center for Population and Animal Health by means of ICP-MS.b The ICP-MS was equipped with an automated samplerc and glass concentric nebulizer.d The instrument was calibrated with appropriate dilutions (1:2 to 1:200,000) of stock (1,000 μg/mL) solutions of each element of interest (arsenic, copper, manganese, selenium, cadmium, iron, mercury, thallium, cobalt, lead, molybdenum, and zinc)e according to in-house–generated standard operating procedures derived from the ICP-MS operator's manual.25 Each quantitative analysis included 2 levels (levels 2 and 3, which contained multiple metals at various concentrations) of a human whole-blood–based metals standard reference materialf for matrix-matched quality control. All results below the 2-ppb limit of quantitation were reported as < 2 ppb.
Data and statistical analysis
Selected RBC morphological changes (basophilic stippling, poikilocytosis, and nucleated RBC count > 100 cells/μL) considered to be potentially consistent with lead exposure were recorded as present or absent for each dog. On the basis of questionnaire review, each test population dog was assigned to one of the following clinical status groups for comparative purposes: no clinical abnormalities, clinical abnormalities considered inconsistent with lead toxicosis, and clinical abnormalities potentially consistent with lead toxicosis. Clinical abnormalities that could be consistent with lead toxicosis included any gastrointestinal or neurologic abnormalities, alone or in combination with other clinical signs. Specifically, dogs reported to have any of the following signs were included in this group: vomiting, diarrhea, anorexia, weight loss, excessive vocalization, behavioral changes (eg, signs of anxiety), and seizures. Reported clinical signs that were not considered to be consistent with lead toxicosis included dermatopathies, arthropathies, ophthalmic abnormalities, and cardiorespiratory abnormalities. Frequency and severity of clinical signs could not be determined in most cases and were not considered in classification. Drinking-water sources were categorized as primary city water source (ie, unfiltered tap water) and alternate water source. Alternate water sources included bottled water, noncity water, and city water filtered for heavy metal removal. Dogs were classified according to the principal source of water as indicated on the questionnaire; responses indicating both primary and alternate water being used as principal sources were excluded from analysis. Dogs were further categorized into those recently drinking primary city water (≤ 3 months) and those not recently (> 3 months) drinking primary city water. Although the elimination half-life of lead in canine blood and exchange rates of blood lead with other compartments are complex and incompletely characterized,26 a time span of 3 months was selected because this would allow for turnover of the majority of RBCs in circulation.27
Blood lead concentration data did not follow a normal distribution as assessed by Kolmogorov-Smirnov testing and box-plot analysis and were reported as median and IQR. For statistical purposes, any BLC < 2 ppb was treated as 1 ppb. Statistical analyses were conducted to compare BLCs of dogs from the test and control populations and to compare BLCs among young (≤ 2 years), middle-age (3 to 5 years), and old (≥ 6 years) dogs within each population. Analyses of test population data were also performed to compare BLCs of dogs with and without clinical signs potentially associated with lead poisoning, to determine whether the water source and duration of time consuming a particular water source was associated with BLCs, and to determine whether BLCs were associated with screening date or city ward of residence. The BLCs from wards 2, 5, and 6 were subsequently grouped together and compared with the values from all remaining wards and regions combined because wards 2, 5, and 6 had been identified as comprising the greatest proportion of households with high lead concentrations in the drinking water.28 Because the distribution of the BLC data was continuous and the observations were independent, the Kruskal-Wallis test and Wilcoxon rank sum test were used for the aforementioned analyses. In addition, proportionate differences in the presence or absence of RBC morphological changes were compared between the test and control populations with the Fisher exact test. Statistical analyses were performed with commercially available software,g and for all analyses, values of P < 0.05 were considered significant.
Intensity maps of Flint depicting residential WLCs and BLCs in dogs were made through the use of commercially available software.h The BLC of each dog in the test population was mapped to the city ward in which the dog resided according to the address provided by the owner. The percentage of dogs within each ward of Flint with BLCs > 5 ppb (a value exceeding the IQR of BLCs for the control group) was calculated with the software provided. The WLCs from 12,161 households were similarly mapped according to water testing data for the period of September 3, 2015, to June 30, 2016, obtained from the Michigan Department of Environmental Quality.28 In this map, the percentages of households within each ward with WLCs > 15 ppb (a cutoff derived from the US Environmental Protection Agency lead and copper rule29 pertaining to the maximum allowable amount of lead in residential tap water) were depicted. The maps were then visually inspected to determine whether areas in which WLCs were > 15 ppb corresponded to areas where dogs with BLCs > 5 ppb resided.
Results
Dogs
Overall, 331 dogs (284 and 47 in the test and control populations, respectively) were included in the study. The median number of dogs tested at the 6 screening clinics was 47 (range, 24 to 72). The 284 test population dogs included 121 females (71 spayed and 50 sexually intact), 135 males (84 neutered and 51 sexually intact), and 28 dogs for which sex and reproductive status were not recorded. The median age of test population dogs was 5 years (IQR, 2.5 to 8 years).
Forty-seven dogs that underwent blood sample collection at the Michigan State University College of Veterinary Medicine during the study period were enrolled in the control population. These included 22 females (17 spayed and 5 sexually intact) and 25 males (16 neutered and 9 sexually intact). The median age of control population dogs was 5 years (IQR, 1.8 to 7.5 years).
BLCs
The median BLC of the test population (4 ppb; IQR, 1 to 8 ppb) was significantly (P = 0.003) greater than that of the control population (1 ppb; IQR, 1 to 4 ppb). Four of 284 (1.4%) dogs in the test population had BLCs > 50 ppb (the reportable limit in the state of Michigan30); 20 had BLCs between 20 and 50 ppb that did not exceed the reportable limit per state of Michigan guidelines, but were concentrations that have been associated with intellectual impairments in children.31 None of the 47 control population dogs had BLCs > 50 ppb, and only 1 had a BLC between 20 and 50 ppb (31 ppb). Overall, 24 (8.5%) test population dogs and 1 (2.1%) control population dog had BLCs ≥ 20 ppb.
Intensity maps of household WLCs and BLCs in test population dogs are provided (Figure 1). No differences in BLCs were detected among the 9 city wards of Flint (P = 0.297); however, when combined BLCs of dogs from wards where the highest WLCs were known to have occurred (2 [n = 15], 5 [19], and 6 [32]) were compared with those of dogs from remaining areas (n = 13, 20, 34, 42, 52, and 24 for wards 1, 3, 4, 7, 8, and 9, respectively; n = 33 for regions outside wards), the values were significantly (P = 0.011) greater for dogs in the former group (5 ppb; IQR, 2 to 9 ppb) than the latter group (3 ppb; IQR, 1 to 7 ppb). The BLCs in the test population at the first screening clinic (held in the western region of Flint) were significantly (P < 0.001) lower than the BLCs at the remaining clinics. No additional differences (P = 0.1) in BLCs were detected among screening clinic dates. Of dogs screened at the initial clinic, 10 of 70 (14.3%) were from the 3 wards (2, 5, and 6) with the greatest proportion of households with high WLCs, whereas 56 of 158 dogs (35.4%) screened at the other clinics were from these wards; further analysis of these findings was not performed.
Intensity maps of 9 Flint, Mich, city wards depicting household WLCs (A) and BLCs in dogs (B) during a public health crisis in which city water was found to be contaminated with lead. A—Map depicting the percentage of households (total, 12,161) in each ward with WLCs > 15 ppb between September 3, 2015, and June 30, 2016.28 B—Map depicting the percentage of dogs (test population; total, 284) residing in each ward that had BLCs > 5 ppb at the time of canine lead-screening clinics (February 20 to May 14, 2016). Cutoffs of 15 and 5 ppb were derived from the US Environmental Protection Agency lead and copper rule for WLCs29 and the value exceeding the upper limit of the IQR of BLCs for a control population of 47 study dogs residing in the East Lansing, Mich, area, respectively. City wards 2, 5, and 6, which had the highest percentages of households with increased WLCs, also had the highest percentages of dogs with BLCs > 5 ppb. The BLCs of dogs in wards 2, 5, and 6 combined were greater than those of dogs in the remaining areas combined (P = 0.011).
Citation: Journal of the American Veterinary Medical Association 251, 8; 10.2460/javma.251.8.912
Further assessment of BLC data for the test population revealed that young dogs (≤ 2 years of age; n = 71) in this group had significantly (P = 0.002) greater BLCs than did old dogs (≥ 6 years of age; 116; median, 5 vs 2.5 ppb, respectively; Figure 2); results did not differ significantly (P = 0.08) in other age group comparisons. The BLC results also did not differ significantly (P = 0.458) among age groups in the control population.
Box-and-whisker plots depicting BLCs of test population dogs (striped boxes) grouped by age (≤ 2 years [n = 71], 2 to 6 years [131], and ≥ 6 years [116]). The BLCs of control population dogs (white box [n = 47]) are provided for reference. The horizontal line within each box represents the median, the lower and upper boundaries of each box represent the first and third quartiles, and the whiskers represent the range. Dots represent the 5% of dogs in each group with the highest BLCs. *The BLCs for dogs ≤ 2 years of age were significantly (P = 0.002) greater than those for dogs ≥ 6 years of age. See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 251, 8; 10.2460/javma.251.8.912
Owner-reported clinical signs potentially consistent with lead toxicosis were recorded for 63 dogs; 34 dogs were reported to have clinical abnormalities considered inconsistent with lead toxicosis, and 187 dogs had no abnormalities reported. The BLCs did not differ significantly (P = 0.218) among these 3 groups.
The source of drinking water was known for 264 of 284 (93.0%) dogs in the test population. At the time of blood sample collection, 184 of 264 (69.7%) dogs were receiving bottled water, 39 (14.8%) were receiving filtered water, 22 (8.3%) were receiving a combination of bottled and filtered water, and 19 (7.2%) were receiving primary city (ie, unfiltered tap) water. No dogs were reported to be receiving a combination of primary city water and water from alternate sources. For the 245 test population dogs receiving water from alternate sources, owners indicated that tap water had been used previously, and changes in drinking water sources were made at various time points following the city's transition to use of water from the Flint River. However, the date of the drinking water source change was available for only 95 dogs. Dogs receiving water from an alternate source for ≤ 3 months had significantly (P = 0.037) greater BLCs (median, 7 ppb; IQR, 3 to 13 ppb; n = 41) than did dogs receiving water from an alternate source for > 3 months (median, 3.5 ppb; IQR, 1 to 7 ppb; 54).
Sources of drinking water were known for 42 control dogs; 4 (9.5%) were receiving filtered water, and 38 (90.5%) were receiving tap water. The drinking water source had not been recently changed (ie, within 12 months) for any of these control population dogs.
Examination of blood smears revealed that 12 of 281 (4.3%) dogs in the test population and 2 of 41 (4.9%) control dogs had high nucleated RBC counts (inadequate sample volume or quality precluded analysis in some dogs). No dogs in either population had evidence of basophilic stippling of erythrocytes. The proportion of dogs with erythrocyte morphology abnormalities was not different between groups (P = 0.695). Hematocrit and presence of poikilocytosis were not compared between groups because artifactual swelling of erythrocytes had resulted from delayed processing of some blood samples from the test population.
Characteristics of dogs with BLCs > 50 ppb
Three of the 4 dogs with BLCs > 50 ppb were from the same household in ward 7 and had BLCs of 121, 77, and 68 ppb (a 2-year-old sexually intact male, a 4-year-old sexually intact female, and a 9-month-old sexually intact male, respectively). All 3 dogs were Australian Shepherds. The dog with the highest BLC had been identified as having lead poisoning by its local veterinarian approximately 3 months prior to BLC measurement for study purposes; at the time of diagnosis, the dog's BLC was 450 ppb. At that time, all 3 dogs in the household were described as having clinical abnormalities including skittish behavior, abnormal and excessive barking, and weight loss; however, the other dogs in the household did not have BLCs tested prior to the lead-screening clinics. The dog known to have lead poisoning was reported to be the most severely affected and also had circling behavior and signs of discomfort initially. The water source for all dogs had been switched to bottled water, and the most affected dog had received chelation therapy with dimercaptosuccinic acid (10.4 mg/kg [4.7 mg/lb], PO, q 8 h) for 10 days. At the time of participation in the lead-screening clinic, clinical signs were reported to be improved but still present in these 3 dogs. The fourth dog with a BLC > 50 ppb (measured value, 58 ppb) was a 6-year-old castrated male mixed-breed dog residing in ward 8 that had no gastrointestinal or neurologic abnormalities reported. Other than recommending a switch to bottled water for drinking, specific treatment was not pursued for this dog. Potential exposure sources were not investigated as a part of this study; however, site investigations were conducted at both households by government agencies. Briefly, drinking water for the 3 dogs in ward 7 was confirmed to be contaminated with lead at concentrations not suitable for consumption by people or animals (> 1,400 ppb in some samples). A site investigation at the home of the dog in ward 8 did not reveal a high WLC in any water source tested, and the source of exposure could not be definitively determined for this dog.
Discussion
The results of the present study suggested that dogs residing in the immediate Flint, Mich, area (ie, the test population) had increased lead exposure, most likely as a result of lead contamination of the city's water supply. Among 95 dogs in the test population for which the date of drinking water source changes were known, those that had been receiving water from an alternate source (ie, other than primary city [unfiltered tap] water) for ≤ 3 months had significantly (P = 0.037) higher BLCs, compared with results for dogs that had been receiving water from an alternate source for > 3 months. Even though only 19 of 264 (7.2%) dogs in the test population were receiving primary city water at the time of study participation, the median BLC of test population dogs (4 ppb; n = 284) was 4 times that of the control population comprising study dogs residing in East Lansing, Mich (1 ppb; 47), and BLCs overall were significantly (P = 0.003) higher in the former group than in the latter. Under normal circumstances, the sole source of drinking water for pet dogs in most households was tap water as evidenced by questionnaire responses for both the control and test populations. This was also likely the predominant overall source of fluid intake for most dogs, given the commonplace practice of feeding dry dog food.32 However, human intake of household tap water is much more variable owing to consumption of commercially prepared beverages (eg, juice, milk, or soda) or food with high moisture content (eg, produce) and the common use of bottled water inside the home, at work, or in public places. As such, we considered it likely that many dogs were initially at even greater risk of lead exposure than were people residing in the same households. This would not be unusual, as dogs can often be sentinels for environmental toxicants.19,20 Unfortunately, minimal relief efforts were aimed at companion animal health during the early periods of this public health crisis. The events in Flint highlight the importance of the one-health approach to raise awareness of the link between human, animal, and environmental health and promote the well-being of all species.21 It should also be noted that these events were not likely to represent an isolated occurrence in the United States. Other postindustrial cities commonly use water lines built with lead-containing pipes or pipes joined with lead-based solder, and substantial numbers of households are speculated to have high WLCs.33 More than 25 water systems in the United States had WLCs > 200 ppb, and thousands of additional water systems in the United States were in violation of the Environmental Protection Agency lead and copper rule29 for failure to appropriately monitor WLCs in 2015.34 Additionally, even strict adherence to the current monitoring guidelines could result in failure to identify some affected households.34 As such, the relevance of the events and findings described in the present study extend beyond the Flint community, and veterinarians are in a unique position to intervene in this potentially far-reaching public health problem.
A noteworthy result of the present study was significantly (P = 0.002) greater BLCs in young dogs (≤ 2 years of age) than in old dogs (≥ 6 years of age) of the test population, a finding that was not observed in control population dogs. It is well established that adolescent people and animals are at greatest risk for lead exposure and its adverse effects. This is attributable to a combination of factors, including behavioral tendencies, greater influx of lead into developing bones (because lead mimics calcium), and greater gastrointestinal absorption of lead during growth and development.1,10,35,36 As such, the age-related difference in BLCs for test population dogs was not unexpected and provided further support for the finding that dogs of this population were impacted by lead exposure through the contaminated water supply. Although no differences were detected in BLCs of test population dogs grouped by the presence or absence of clinical signs, 3 of 4 dogs with BLCs > 50 ppb had owner-reported gastrointestinal or neurologic abnormalities at the time of screening, despite the fact that BLCs were considered only mildly to moderately increased. Clinical signs of lead toxicosis in people and animals historically were thought to occur at much greater BLCs, often at concentrations > 350 or 400 ppb.1–5,10 Conversely, more recent reports10,31,37,38 concerning human lead exposure have documented that low-level exposures associated with much lower BLCs can still result in a plethora of cognitive, developmental, and behavioral abnormalities that are apparent on a population basis, but not readily identifiable on an individual basis. Indeed, subclinical BLC abnormalities associated with chronic low-level lead exposures may render dogs more susceptible to atypical adverse effects, such as alterations in immune system responsiveness, cognition, or organ function. The observations in our study raised the possibility that some clinical signs in these dogs may have been related to previous or ongoing lead exposure; however, this could not be confirmed, as other demographic factors or disease processes could have been responsible for these observations, and evaluation of possible associations between BLCs and various clinical signs did not reveal significant findings. Nonetheless, additional investigation in this area is warranted.
Measurement of BLC is a widely used diagnostic test for lead exposure in both human and veterinary medicine because circulating RBCs rapidly take up lead following its gastrointestinal absorption.1–5,10,12,15 Given the elimination half-life of lead bound to RBCs, the optimal window for BLC testing is during ongoing exposure or within several weeks after exposure.36 However, lead is widely distributed to other tissues, where it can accumulate and induce pathological changes. The elimination of lead from some tissues, such as bone or CNS tissues, is extremely prolonged.26,36 Lead that has accumulated in these storage sites can gradually redistribute into the peripheral circulation, but BLCs would not necessarily correspond to the overall lead burden.26 Consequently, an animal with previous exposure or with chronic low-level exposure can have a high lead burden despite a low BLC. In keeping with this premise, a recent report described pathological bone fractures in a dog with markedly increased bone lead concentrations but BLC within the laboratory reference range.14 Considering this information, it would be expected that more dogs in our test population were affected by lead exposure than the data suggested. In October 2015 (4 months prior to the first lead-testing clinic in this study), the city of Flint resumed use of water from Lake Huron, and in December of 2015, the city added additional corrosion inhibitors to the water supply, measures that lessened household WLCs.39,40 Additionally, most (245/264 [92.8%]) test population dogs were no longer being provided unfiltered tap water at the time of BLC measurements, thus further minimizing their ongoing lead exposure. Beyond the lead-screening clinics, the Michigan State University College of Veterinary Medicine convened a Flint response team that included members of the Michigan Department of Agriculture and Rural Development. The mission of this team was not limited to an epidemiological investigation. An initial priority was to provide educational assistance to the Flint community to help raise awareness of the issue and provide animal owners with information about lead poisoning. This was largely accomplished through interviews with representatives from various news outlets and by the generation and distribution of flyers at water distribution sites as well as local churches and other community organizations. Also, members of the Michigan Veterinary Medical Association received an update letter with information on the medical and regulatory aspects of lead toxicosis. Although no livestock were found to be consuming water from the city of Flint, the initial priority of the Michigan Department of Agriculture and Rural Development was to assess for the presence of livestock in the affected area to address potential food safety concerns. Collectively, these measures delayed the onset of BLC screening for dogs in Flint. Still, 4 dogs in the test population of the present study had BLCs > 50 ppb; an additional 20 had BLCs > 20 ppb, and overall, BLCs were significantly greater in the test population than in the control population.
Multiple strategies for obtaining blood samples from test population dogs were considered. Given the sensitive political environment and community member distrust in government agencies, the investigators believed it was important to work with local community organizations that had well-established relationships with Flint residents.18,23,24 Partnership with local veterinary clinics was considered as an option, but this was impractical because there were few veterinarians working within affected city wards. Furthermore, samples obtained from veterinary clinics could select for a population of dogs receiving routine veterinary care, which might not be an accurate representation of the canine population. Public schools also were considered, but government health and safety regulations precluded their use. Under the circumstances, we sought to engage well-established community organizations such as animal shelters and religious centers. The 5 organizations that allowed use of their facilities for the clinics had been active in ongoing relief efforts such as public education and water distribution. In addition, these groups had a strong desire to assist in relief efforts that included companion animal health. The range of participation at individual clinics was variable (24 to 72 dogs), but subjectively, clinics were well received by the public. Owners of participating dogs appeared to welcome the initiative and to be grateful for the opportunity to test their dogs for lead exposure.
One limitation of the present study was its small sample size. The investigators attempted to promote clinics through various media, but the success of those strategies in reaching members of the urban community, particularly in more disadvantaged areas, was unknown. Some organizations were reluctant to disseminate or publish promotional material. Furthermore, some residents may not have perceived a need to test dogs that had no apparent clinical signs, especially if they were receiving bottled water. Inclement weather (eg, a snowstorm during 1 clinic) may have resulted in low attendance as well. Whether these or other more complex and interacting factors influenced participation was unknown. Even though participant numbers were limited, informative and valuable results were obtained. Another limitation was the potential for recall bias, especially as it related to owner-reported clinical abnormalities. Veterinarians did not perform health assessments or review medical records of participating dogs, and owner responses could have been influenced by a heightened regional awareness of lead-induced illness. These limitations might have affected the investigation of potential associations between BLCs and clinical abnormalities. The lack of a random sampling strategy was also a potential source of bias; however, its impact could not be estimated. For instance, dogs from high-risk, underprivileged areas could have been underrepresented because of lack of transportation, poor dissemination of information, or weekend work requirements. Conversely, this population could have been overrepresented, given the known exposure risk to people in some of these neighborhoods.15
It should also be noted that the cross-sectional nature of the present study did not allow us to definitively conclude that BLC abnormalities observed in the present report were a direct result of the water source change to the Flint River. High BLCs are most often observed in underprivileged inner-city neighborhoods, and this is likely related to aging infrastructure (eg, older pipes and paint) in such areas.41,42 Socioeconomic differences were known to exist between Flint and East Lansing,43,44 and it is possible that some differences in canine BLCs existed prior to the water crisis. Aged household paint, which remains an important source of exposure for children and livestock in some regions, or other nonwater sources of environmental lead contamination could have contributed to the BLC findings in dogs.45–47 Ideally, a longitudinal study or a study comparing canine BLC data before and after the water source change would have been performed. Unfortunately, BLC data available for dogs in Michigan at any time point are scarce. However, the results of numerous human studies48–54 suggest that lead from drinking water is an important source of detectable circulating lead concentrations in children48–51 and that water corrosiveness influences lead leaching and WLCs.52–54 Given that Flint River water was more corrosive to the city's aging lead plumbing, which led to increased WLCs and BLCs in children,15,17 we believe it likely that the change in water source contributed to lead exposure in the test population of dogs. The findings reported herein, including the differences in BLCs between the test and control populations, the differences in BLCs between young and old dogs within the test population, and the differences in BLCs detected according to the duration of time dogs had been receiving water from alternate sources, all supported the role of city water in the BLC abnormalities identified.
Lead poisoning is likely an underrecognized condition in companion animals in the United States, and the events described here highlight the importance of veterinary clinicians maintaining an awareness of environmental contaminants such as lead. Veterinarians should also remain cognizant that lead exposure and other chemically induced toxicoses are reportable diseases in some regions. Ideally, veterinarians should be included in early response efforts in emergencies affecting water supplies, given the likelihood of impacts on animal health. Furthermore, these events draw attention to the importance of a one-health approach in which both human and animal well-being are considered in a public health crisis.
Acknowledgments
Supported by The Hazel and Melvin Cant Companion Animal Endowed Research Fund of the Michigan State University College of Veterinary Medicine, the Companion Animal Gift Fund of the Michigan State University College of Veterinary Medicine, and the Michigan Department of Agriculture and Rural Development. Ms. Daniels was supported in part by NIH grant 4T35OD016477-14 to Michigan State University.
The authors declare that there were no conflicts of interest.
Presented in part in poster form at the Merial-NIH National Veterinary Scholars Forum, Columbus, Ohio, August 2016.
ABBREVIATIONS
| BLC | Blood lead concentration |
| ICP-MS | Inductively coupled plasma–mass spectrometry |
| IQR | Interquartile range |
| ppb | Parts per billion |
| WLC | Water lead concentration |
Footnotes
Advia 120 Hematology System, Siemens Healthcare, Deerfield, Ill.
Agilent 7500ce ICP-MS, Agilent Technologies, Santa Clara, Calif.
Cetac autosampler, Cetac, Omaha, Neb.
MicroMist nebulizer, Agilent Technologies, Santa Clara, Calif.
Specpure, Alfa Aesar, Ward Hill, Mass.
Lyphochek Whole Blood Metals standard reference material, Bio-Rad Laboratories, Hercules, Calif.
SAS, version 9.1, SAS Institute Inc, Cary, NC.
ArcGIS Desktop, Environmental Systems Research Institute Inc, Redlands, Calif.
References
1. Kowalczyk DF. Lead poisoning in dogs at the University of Pennsylvania Veterinary Hospital. J Am Vet Med Assoc 1976; 168: 428–432.
2. Prescott CW. Clinical findings in dogs and cats with lead poisoning. Aust Vet J 1983; 60: 270–271.
3. Zook BC, Carpenter JL, Leeds EB. Lead poisoning in dogs. J Am Vet Med Assoc 1969; 155: 1329–1342.
4. McLeavey BJ. Lead poisoning in dogs. N Z Vet J 1977; 25: 395–396.
5. Morgan RV, Moore FM, Pearce LK, et al. Clinical and laboratory findings in small companion animals with lead poisoning: 347 cases (1977–1986). J Am Vet Med Assoc 1991; 199: 93–97.
6. Morgan RV. Lead poisoning in small companion animals: an update (1987–1992). Vet Hum Toxicol 1994; 36: 18–22.
7. Penumarthy L, Oehme FW, Galitzer SJ. Effects of chronic oral lead administration in young Beagle dogs. J Environ Pathol Toxicol 1980; 3: 465–490.
8. Zook BC, Carpenter JL, Roberts RM. Lead poisoning in dogs: occurrence, source, clinical pathology, and electroencephalography. Am J Vet Res 1972; 33: 891–902.
9. Johnson BL, Mason RW. A review of public health regulations on lead. Neurotoxicology 1984; 5: 1–22.
10. Needleman H. Lead poisoning. Annu Rev Med 2004; 55: 209–222.
11. Hamir AN. Review of lead poisoning in dogs. Vet Bull 1986; 56: 1059–1070.
12. Jones RL, Homa DM, Meyer PA, et al. Trends in blood lead levels and blood lead testing among US children aged 1 to 5 years, 1988–2004. Pediatrics 2009; 123: e376–e385.
13. Bischoff K, Priest H, Mount-Long A. Animals as sentinels for lead exposure: a case report. J Med Toxicol 2010; 6: 185–189.
14. Cole G, Weigel J, Headrick A, et al. Multiple pathologic fractures and delayed union associated with lead exposure in a German Shepherd Dog. Aust Vet J 2015; 93: 373–376.
15. Hanna-Attisha M, LaChance J, Sadler RC, et al. Elevated blood lead levels in children associated with the Flint drinking water crisis: a spatial analysis of risk and public health response. Am J Public Health 2016; 106: 283–290.
16. Southall A. State of emergency declared over man-made water disaster in Michigan city. New York Times2016;Jan 17. Available at: www.nytimes.com/2016/01/18/us/obama-flint-michigan-water-fema-emergency-disaster.html. Accessed May 30, 2017.
17. Edwards M, Falkinham J, Pruden A. Synergistic impacts of corrosive water and interrupted corrosion control on chemical/microbiological water quality: Flint, MI. National Science Foundation Grant abstract. Available at: www.nsf.gov/award-search/showAward?AWD_ID=1556258&HistoricalAwards=false. Accessed May 30, 2017.
18. Gostin LO. Politics and public health: the Flint drinking water crisis. Hastings Cent Rep 2016; 46: 5–6.
19. Backer LC, Grindem CB, Corbett WT, et al. Pet dogs as sentinels for environmental contamination. Sci Total Environ 2001; 274: 161–169.
20. Buck WB. Animals as monitors of environmental quality. Vet Hum Toxicol 1979; 21: 277–284.
21. King LJ, Anderson LR, Blackmore CG, et al. Executive summary of the AVMA One Health Initiative Task Force report. J Am Vet Med Assoc 2008; 233: 259–261.
22. East Lansing–Meridian Water and Sewer Authority. 2015 drinking water quality report for the city of East Lansing & Meridian Township. Available at: www.cityofeastlansing.com/ArchiveCenter/ViewFile/Item/437. Accessed May 30, 2017.
23. Dixon J. How Flint's water crisis unfolded. Detroit Free Press 2016;Jan 24. Available at: www.freep.com/pages/interactives/flint-water-crisis-timeline/. Accessed May 30, 2017.
24. Bosman J. After water fiasco, trust of officials is in short supply in Flint. New York Times 2016;Oct 8. Available at: www.nytimes.com/2016/10/09/us/after-water-fiasco-trust-of-officials-is-in-short-supply-in-flint.html. Accessed May 30, 2017.
25. Agilent 7500 Series ICP-MS ChemStation (G1834B) operator's manual. Santa Clara, Calif: Agilent Technologies, 2005.
26. Abadin H, Ashizawa A, Stevens Y, et al. Toxicological profile for lead. Atlanta: US Department of Health and Human Services, Agency for Toxic Substances and Diseases Registry, 2007. Available at: www.atsdr.cdc.gov/toxprofiles/tp.asp?id=96&tid=22. Accessed May 30, 2017.
27. Novinger MS, Sullivan PS, McDonald TP. Determination of the lifespan of erythrocytes from Greyhounds, using an in vitro biotinylation technique. Am J Vet Res 1996; 57: 739–742.
28. Michigan Department of Environmental Quality. Flint water. Reports/requests. Available at: www.michigan.gov/flintwater/0,6092,7–345–76292_76294_76297—,00.html. Accessed May 30, 2017.
29. United States Environmental Protection Agency. Lead and copper rule. Available at: www.epa.gov/dwreginfo/lead-and-copper-rule. Accessed May 30, 2017.
30. Reportable animal diseases in Michigan. Available at: www.michigan.gov/documents/mdard/Website_update_form_AID_532783_7.pdf. Accessed May 30, 2017.
31. Canfield RL, Henderson CR, Cory-Slechta DA, et al. Intellectual impairments in children with blood lead concentrations below 10 μg per deciliter. N Engl J Med 2003; 348: 1517–1526.
32. Slater MR, Robinson LE, Zoran DL, et al. Diet and exercise patterns in pet dogs. J Am Vet Med Assoc 1995; 207: 186–190.
33. Seville LR. Is Philadelphia testing its drinking water correctly? NBC News 2016;Feb 19. Available at: www.nbcnews.com/storyline/flint-water-crisis/philadelphia-testing-its-drinking-water-correctly-n521036. Accessed May 30, 2017.
34. Ganim S. Over 5,300 US water systems are in violation of lead rules. CNN 2016;Jun 29. Available at: www.cnn.com/2016/06/28/us/epa-lead-in-u-s-water-systems/. Accessed May 30, 2017.
35. Goyer RA, Mahaffey KR. Susceptibility to lead toxicity. Environ Health Perspect 1972; 2: 73–80.
36. Gwaltney-Brant S. Metals and minerals (lead). In: Plumlee KH, ed. Clinical veterinary toxicology. St Louis: Mosby, 2004; 204–210.
37. Blackowicz MJ, Hryhorczuk DO, Rankin KM, et al. The impact of low-level lead toxicity on school performance among Hispanic subgroups in the Chicago public schools. Int J Environ Res Public Health 2016; 13: E774.
38. Lanphear BP, Dietrich K, Auinger P, et al. Cognitive deficits associated with blood lead concentrations < 10 micrograms/dL in US children and adolescents. Public Health Rep 2000; 115: 521–529.
39. Fonger R. Flint officials say they've boosted corrosion control agents in water. MLive 2015;Dec 9. Available at: www.mlive.com/news/flint/index.ssf/2015/12/flint_officials_say_theyre_boo.html#incart_river_home. Accessed May 30, 2017.
40. Kennedy M. Lead-laced water in Flint: a step-by-step look at the makings of a crisis. WKAR (NPR) Public Media 2016;Apr 20. Available at: www.npr.org/sections/thetwo-way/2016/04/20/465545378/lead-laced-water-in-flint-a-step-by-step-look-at-the-makings-of-a-crisis. Accessed May 30, 2017.
41. McLaine P, Navas-Acien A, Lee R, et al. Elevated blood lead levels and reading readiness at the start of kindergarten. Pediatrics 2013; 131: 1081–1089.
42. Moody HA, Darden JT, Pigozzi BW. The relationship of neighborhood socioeconomic differences and racial residential segregation to childhood blood lead levels in metropolitan Detroit. J Urban Health 2016; 93: 820–839.
43. City-Data.com. East Lansing, Michigan. Available at: www.city-data.com/city/East-Lansing-Michigan.html. Accessed May 30, 2017.
44. City-Data.com. Flint, Michigan. Available at: www.city-data.com/city/Flint-Michigan.html. Accessed May 30, 2017.
45. Herman T. Lansing's lead problem. CityPulse 2016;Feb 17. Available at: lansingcitypulse.com/article-12770-lansing's-lead-problem.html. Accessed May 30, 2017.
46. Jacobs DE, Clickner RP, Zhou JY, et al. The prevalence of lead-based paint hazards in US housing. Environ Health Perspect 2002; 110: A599–A606.
47. Thompson L. Lead. In: Gupta R, ed. Veterinary toxicology: basic and clinical principles. 2nd ed. New York: Elsevier Academic Press, 2012; 522–526.
48. Brown MJ, Margolis S. Lead in drinking water and human blood lead levels in the United States. MMWR Suppl 2012; s61: 1–9.
49. Ngueta G, Abdous B, Tardif R, et al. Use of a cumulative exposure index to estimate the impact of tap-water lead concentration on blood lead levels in 1- to 5-year-old children (Montreal, Canada). Environ Health Perspect 2016; 124: 388–395.
50. Triantafyllidou S, Edwards M. Lead (Pb) in tap water and in blood: implications for lead exposure in the United States. Crit Rev Environ Sci Technol 2012; 42: 1297–1352.
51. Triantafyllidou S, Gallagher D, Edwards M. Assessing risk with increasingly stringent public health goals: the case of water lead and blood lead in children. J Water Health 2014; 12: 57–68.
52. Edwards M, McNeill LS. Effect of phosphate inhibitors on lead release from pipes. J Am Water Works Assoc 2002; 94: 79–90.
53. Edwards M, Triantafyllidou S. Chloride-to-sulfate mass ratio and lead leaching to water. J Am Water Works Assoc 2007; 99: 96–109.
54. Miranda ML, Kim D, Hull AP, et al. Changes in blood lead levels associated with use of chloramines in water treatment systems. Environ Health Perspect 2007; 115: 221–225.
