Turtles were inexpensive, popular, and low maintenance, with an array of groovy pre-molded plastic housing designs to choose from. Invariably they would escape, only to be found days later behind the couch along with the skeleton of the class bunny my younger sister brought home from kindergarten one weekend.
But eventually, replacement turtles became harder to come by. Reports started surfacing that people with pet turtles were getting sick. In 1975, the U.S. Food and Drug Administration (FDA) banned commercial distribution of turtles less than 4 inches in length, and it has been estimated that the FDA ban prevents some 100,000 cases of salmonellosis among children each year.
Maybe I picked up my turtle, rolled around on the carpet with it, pet it a bit, and then stuck my finger in my mouth. Maybe in my emotionally vacant adolescence I kissed my turtle. Who can remember?
The U.S. Centers for Disease Control reports today that in June 2017, PulseNet, the national molecular subtyping network for foodborne disease surveillance, identified 17 Salmonella Agbeni clinical isolates with indistinguishable XbaI enzyme pattern (outbreak strain) by pulsed-field gel electrophoresis. The same Salmonella Agbeni XbaI pattern was isolated from a turtle in 2015; in a 2016 investigation involving the same outbreak strain, 63% of patients reported contact with turtles (CDC, unpublished data, 2016). Despite prohibition of sale of small turtles (shell length less <4 inches) in the United States since 1975 (1), illness outbreaks associated with turtle contact continue to occur. Ill persons in previous Salmonella Poona and Salmonella Pomona outbreaks linked to turtles were geographically concentrated in the Southwest region of the United States (2,3). Turtle production is known to be higher in the Southeast region of the country (2). An outbreak investigation by CDC and health departments was initiated to identify the source of the 2017 illness outbreak.
A case was defined as isolation of Salmonella Agbeni with the outbreak strain from an ill patient during April–December 2017. State and local health officials interviewed patients to ascertain turtle exposure information, including details about the species of turtle and purchasing information. Purchase locations reported by patients were contacted for traceback information. Whole genome sequencing (WGS), using high quality single nucleotide polymorphism (hqSNP) analysis, was performed by CDC on clinical isolates from the 2017 outbreak, the 2016 illness cluster, and the turtle isolate from 2015 to characterize genetic relatedness.
Seventy-six cases were identified in 19 states in 2017; two thirds (67%) of patients resided in East Coast states (Connecticut, Delaware, Maryland, Massachusetts, New Jersey, New York, Pennsylvania, Rhode Island, and Virginia).* Patient ages ranged from <1–100 years (median = 21 years). Among 63 (83%) patients with information on hospitalization, 30 (48%) were hospitalized; no deaths were reported. Fifty-nine (78%) patients provided exposure information, including 23 (39%) who reported contact with turtles; among these, 14 (61%) specified small turtles. Among 12 patients who reported how the turtles were obtained, six purchased them from a street or roadside vendor, three purchased them from a retail store, two purchased them at festivals, and one reported receiving them as a gift. The traceback investigation did not identify a common turtle farm that supplied purchase locations. WGS hqSNP analysis indicated that the 2017 and 2016 clinical isolates and the 2015 turtle isolate were closely related, differing by 0–18 SNPs.
This salmonellosis outbreak was linked to contact with small turtles and was associated with a higher frequency of hospitalization (48%) than multistate foodborne pathogen outbreaks (27%) as well as recent Salmonella outbreaks linked to turtles (28%–33%) (2–4). The geographic distribution of patients differed from that of previous outbreaks, suggesting the need to better understand the breeding of turtles and distribution of turtle sales in the United States. WGS hqSNP analysis was used to link historic illnesses and turtle isolates to isolates from 2017 patients, supporting the hypothesis that turtles were the likely source of this outbreak. This outbreak indicates further need to educate consumers and retail store staff members regarding the ban on sale of small turtles and to educate consumers to prevent transmission of Salmonella from pets to humans.
Notes from the Field: An Outbreak of Salmonella Agbeni Infections Linked to Turtle Exposure — United States, 2017
Lia Koski, MPH1,2; Lauren Stevenson, MHS1,3; Jasmine Huffman1; Amy Robbins, MPH4; Julia Latash, MPH5,6; Enoma Omoregie, PhD5; Kelly Kline, MPH7; Megin Nichols, DVM1 (View author affiliations)
Gambino-Shirley K, Stevenson L, Concepción-Acevedo J, et al. Flea market finds and global exports: four multistate outbreaks of human Salmonellainfections linked to small turtles, United States—2015. Zoonoses Public Health 2018;65:560–8. CrossRefPubMed
Basler C, Bottichio L, Higa J, Prado B, Wong M, Bosch S. Multistate outbreak of human SalmonellaPoona infections associated with pet turtle exposure—United States, 2014. MMWR Morb Mortal Wkly Rep 2015;64:804. CrossRefPubMed
Health inspectors responded to a complaint about the Good Fortune Supermarket on Tuesday and found 45 frogs and seven turtles for sale at the establishment, said Joseph Wendelken, spokesman for the Rhode Island Department of Health.
“It’s against Rhode Island food code for there to be live animals in a food establishment,” he said.
The department ordered the owners to stop selling the animals and put up signs at the store to notify the public that the critters were not for sale, Wendelken said.
Health officials will coordinate with the Rhode Island Department of Environmental Management to determine what to do with the animals, he said. For now, they are still at the store.
Good Fortune, which is described on its website as one of the largest Asian supermarkets in the United States, opened in Providence about two weeks ago, Wendelken said. The market has more than a dozen locations across the U.S., according to its website.
Health officials inspected the establishment before it opened and found no violations, Wendelken said.
“There was no indication at that point that frogs and turtles were going to be sold,” he said.
Following an impassioned speech by Australian deputy PM Barnaby Joyce (right, not exactly as shown) on government plans to deliberately infect invasive carp with herpes, an increasing number of green sea turtles on Australia’s Great Barrier Reef, with pollution being investigated as the prime culprit.
The animals have a turtle-specific herpesvirus that causes fibropapillomatosis – a condition in which disfiguring tumours grow on the eyes, flippers, tail, shell or internal organs.
“The tumours are benign but can grow up to 30 centimetres in size and block the turtles’ vision, says Karina Jones of James Cook University in Townsville, Australia. “This means they can’t find food or see predators or boats.”
Turtles with tumours are also more vulnerable to other infections, she says. “Severely affected turtles are quite skinny and have other pathogens affecting them – that’s why they die.”
The unpublished results of surveys by Jones’s team this year show that herpesvirus is most prevalent within a narrow stretch of Cockle Bay at Magnetic Island, a popular tourist destination in the middle of the reef. Roughly half the turtles in this hotspot have fibropapillomatosis, compared with less than 10 per cent of turtles sampled across the rest of Cockle Bay.
The cause remains unclear, but environmental contaminants are at the top of the suspect list. “We see these tumours in turtles in very localised hotspots around the world where there is heavy human activity,” says Jones.
In August 2015, the U.S. Food and Drug Administration (FDA) notified the U.S. Center for Disease Control of a consumer complaint involving Salmonella Sandiego infection in a child (the index patient), who had acquired a small turtle (shell length <4 inches [<10 cm]) at an Alabama flea market. The subsequent investigation, which included examining data from PulseNet, the national molecular subtyping network for foodborne disease surveillance, identified four multistate Salmonella outbreaks: two involving Salmonella Sandiego and two involving Salmonella Poona. Theseserotypes have been linked to small turtles in previous outbreaks (1,2).
Although selling small turtles as pets in the United States has been banned since 1975 (3), illegal sales still occur at discount stores and flea markets and by street vendors. CDC investigated to determine the extent of the outbreaks and prevent additional infections.
For this investigation, a case was defined as infection with Salmonella serotypes Sandiego or Poona with the pulse-field gel electrophoresis patterns XbaI JL6X01.0104, JL6X01.0544, JL6X01.0055, JLXX01.0030, or JLXX01.0053 in a person with illness onset January 1–December 31, 2015. A total of 124 cases from 22 states were identified. Median patient age was 7 years (range <1–82 years); 51 patients (41%) were aged <5 years, and 59 (49%) were female. Among 108 patients with available information, 36 (33%) were hospitalized. Thirteen (36%) of the 36 hospitalized patients were aged <5 years. No deaths were reported. Fifty-eight (70%) of 83 patients with ethnicity information were Hispanic.
A total of 104 patients or their caregivers were interviewed; 50 patients (48%) had turtle exposure. Among the 50 patients with reported turtle exposure, the median age was 3 years (range <1–77 years), 30 (60%) were aged <5 years, and 25 (50%) were female. Of the 50 patients with turtle exposure, 48 had information on hospitalization; 18 of the 48 (38%) were hospitalized. Nine of those hospitalized were aged <5 years. Twenty-eight (72%) of 39 patients with ethnicity information were Hispanic. Turtle and turtle environment samples were collected from nine patients’ homes. Of these, 21 isolates of Salmonella were culture-positive: 17 matched the outbreak strains, two were Salmonella Paratyphi B var L-tartrate+, and two were Salmonella Pomona (Table).
An FDA consumer safety officer investigating the flea market where the index patient’s turtle was obtained traced the turtle’s origin to farm A (FDA’s Center for Veterinary Medicine did not pursue legal action regarding the sale by farm A of small turtles to the flea market vendor). Pond water samples from farm A yielded cultures of Salmonella (Table). Although the samples from turtle farm A did not yield the 2015 outbreak strains, two of the three Salmonella Pomona isolates from pond water matched isolates from a patient’s pet turtle’s environment.
The high proportion of patients aged <5 years in these outbreaks emphasizes the risk for illness among young children with small turtles as pets. This finding indicates education is needed to address the risk to children for Salmonella infections from small turtles and other reptiles. Many patients in these outbreaks were Hispanic, which highlights the importance of multilingual messages for effective communication. During these outbreaks, the announcement and prevention messages regarding safe handling of pet reptiles and amphibians were translated into Spanish and posted on the “CDC en Español” website. To ensure wider dissemination, information was also provided to a major Spanish-language news agency.
Transient vendors continue to be a source of illegal sales of small turtles. Educational campaigns at schools, child care centers, carnivals, and flea markets regarding the risk for Salmonella from contact with turtles are important to prevent infections among children.
Sales of turtles with shells less than 4 inches long have been banned in the United States since the 1970s because the creatures are known carriers of Salmonella, which can cause fever, abdominal cramps and diarrhea so severe that some patients need hospitalization.
Despite the sales ban and the known risk, salmonellosis outbreaks tied to turtles have increased since 2006, a research team led by scientists from the Centers for Disease Control and Prevention report in the journal Pediatrics.
“All turtles — healthy and sick, big and small — can carry Salmonella,” said lead author Dr. Maroya Walters, an epidemiologist at the CDC in Atlanta. “Because young children have less developed immune systems and are more likely to engage in hand-to-mouth behaviors, turtles of any size are not appropriate pets for households, schools or daycares with children younger than 5 years of age.”
Salmonella is part of the normal gut flora of turtles, and there’s no way to distinguish a healthy turtle from an infected one, researchers note. The bacteria are present in feces as well as surfaces and water that the animals touch, making it easy for infections to spread to kids who touch the turtles or play with the tank or habitat.
The majority of cases were seen in children younger than 5 years old, and 68% of patients reported a recent exposure to a small turtle.
In the past, exposure to reptiles and amphibians led to thousands of Salmonella infections, particularly in young children. Due to these infections and the potential for developing invasive disease, a federal ban was created in 1975 on the sale and distribution of small turtles less than 4 inches. However, recent multistate outbreaks prompted the current investigative study. Authors of this study sought to understand the epidemiology of turtle-associated salmonellosis and turtle care by case-patients, and to determine where these outbreaks originated. A total of 8 individual, multistate outbreaks were identified. The majority of cases were accounted for by young children, who reported recent exposure to small turtles. A minimal amount of participants were previously aware of the Salmonella reptile association. Investigative efforts traced many cases to 2 turtle farms in Louisiana. After distribution from these farms was halted, the number of new cases of Salmonella infections decreased drastically. Data may be limited as several cases were linked to untraceable sources. Nonetheless, this study should encourage pediatricians to warn their patients and parents about the potential dangers of animal exposure, particularly to small turtles.
In order to identify outbreaks of Salmonella, researchers used data between May 2011 and September 2013 from local and state reference laboratories as well as from the Centers for Disease Control and Prevention. Outbreaks were described as epidemiological and environmental sampling links between ≥2 case-patients including multiple serotypes. Cases were defined as an infection with ≥1 of the outbreak strains. Case-patients were interviewed and completed a questionnaire detailing contact with turtles and their knowledge of the association between reptiles and Salmonella. Swab and water samples were collected by state and local health agencies from patient homes and retail establishments where cases were reported. Researchers identified 8 separate, multi-state Salmonella outbreaks for a total of 473 individual cases. Affected patients <5 years old accounted for 55% of all cases and Hispanic ethnicity was reported in 45% of cases. Turtle exposure was identified in 68% of interviewed patients, and 88% of these were due to contact with small turtles. The association between reptiles and Salmonella was previously known in only 15% of case-patients. Salmonellae were confirmed to be present in 5 retail stores in Florida, which were subsequently traced back to 2 separate Louisiana turtle farms.
“The experience was torturous to us,” said Ms Gibson. “It is impossible to put into words how painful it is to see one’s newborn in that state. It was equally as painful to have to walk away and leave him there night after night.
“I wanted to get rid of the turtles before Oliver was born because I thought they smelled awful and I was worried about salmonella. The only reason we ended up keeping them was because we couldn’t find anyone to take them. I didn’t want to just take them to a pond and dump them so we kept them.”
The infants presented at age 11 days with poor feeding and lethargy, hypotonia, dilated pupils and absent reflexes. Fecal samples were positive for C. butyricum BoNT/E. The infants recovered after treatment including botulism immune globulin intravenous (BIG-IV).
C. butyricum BoNT/E was isolated from water from tanks housing pet ‘yellow-bellied’ terrapins (Trachemys scripta scripta): in case A the terrapins were in the infant’s home; in case B a relative fed the terrapin prior to holding and feeding the infant when both visited another relative. C. butyricum isolates from the infants and the respective terrapin tank waters were indistinguishable by molecular typing. Review of a case of C. butyricum BoNT/E botulism in the UK found that there was a pet terrapin where the infant was living.
It is concluded that the C. butyricum-producing BoNT type E in these cases of infant botulism most likely originated from pet terrapins. These findings reinforce public health advice that reptiles, including terrapins, are not suitable pets for children aged <5 years, and highlight the importance of hand washing after handling these pets.
Infant botulism due to C. butyricum type E toxin: a novel environmental association with pet terrapins
E.B. Shelley, D. O’Rourke, K. Grant, E. McArdle, L. Capra, A. Clarke, E. McNamara, R. Cunney, P. McKeown, C.F.L. Amar, C. Cosgrove, M. Fitzgerald, P. Harrington, P. Garvey, F. Grainger, J. Griffin, B.J. Lynch, G. McGrane, J. Murphy, N. Ni Shuibhne and J. Prosser