A repeated cross-sectional study of intestinal parasites in Texas shelter dogs using fecal flotation and saline sedimentation

Abstract

Estimates of intestinal parasite prevalence in canine populations have largely been based on use of fecal flotation methods only. Dogs in animal shelters are likely at higher risk of intestinal parasite infection because of their previous exposure history. Our objectives were to estimate the prevalence of intestinal parasites among Texas shelter dogs using centrifugal fecal flotation and saline sedimentation techniques, to identify risk factors for infection, and to compare proportions of positive samples detected via fecal flotation vs. saline sedimentation for the most common parasites. Using a repeated cross-sectional study design, we collected fecal samples from dogs on three visits to each of seven Texas animal shelters between May 2013 and December 2014. Fecal flotation and/or saline sedimentation were used to identify parasites in samples. Fecal samples were collected from 529 dogs. The most frequently detected parasites were Ancylostoma caninum (26.4% via fecal flotation, 20.7% via saline sedimentation) and Trichuris vulpis (12.0% via fecal flotation, 14.1% via saline sedimentation). Risk factors for certain parasites were identified; for example, dogs with abnormal fecal consistency were more likely to be shedding T. vulpis eggs than dogs with normal fecal consistency (OR = 1.8, p = 0.005). The addition of fecal sedimentation not only added to the number of parasite species detected in this study, but it also increased the number of dogs diagnosed with the common intestinal parasites that are primarily detected using fecal flotation methods. Texas shelter dogs carry a high burden of intestinal parasites, including those of zoonotic importance.

Introduction

Canine intestinal parasite survey studies vary widely in their prevalence estimates based on factors such as population of dogs sampled (shelter vs. client owned), age ranges sampled, geographic location, and parasitologic tests performed. In most studies, the helminths most frequently reported are Ancylostoma caninum, the common canine hookworm, Toxocara canis, the canine roundworm, and Trichuris vulpis, the canine whipworm. In a US study based on records for more than 1.2 million dogs from a large corporate chain of veterinary hospitals, the prevalence estimates of Ancylostoma sp., Toxocara sp., and Trichuris sp. were 4.5%, 5.0%, and 0.8%, respectively (Mohamed et al. 2009). In a US survey of over 1.1 million canine fecal samples using data from a large commercial fee-for-service diagnostic laboratory system, the prevalence estimates of hookworms, ascarids, and whip-worms were 2.5%, 2.2%, and 1.2%, respectively (Little et al. 2009). Additional parasites reported in that study were Cystoisospora (4.4%) and Giardia (4.0%). The canine patient populations in both of these surveys were likely composed predominantly of pet dogs, as the testing was conducted on samples submitted through veterinary hospitals. However, surveys conducted in canine populations from animal shelters typically yield higher estimates of intestinal parasite prevalence. In a US national shelter dog intestinal parasite prevalence study, in which 6,458 samples were collected from 1993 to 1994, the prevalence estimates of A. caninum, T. canis, and T. vulpis were 19.2%, 14.5%, and 14.3%, respectively (Blagburn et al. 1996). In a second national survey with samples collected at least 15 years later, intestinal parasite prevalence among shelter dogs (n = 6,418) appeared to be similar. An interim report in 2014 showed prevalence estimates of A. caninum, T. canis, and T. vulpis of 29.8%, 12.5%, and 18.6%, respectively, suggesting that the common helminth parasite burdens in stray and shelter dogs are not decreasing over time (Blagburn et al. 2014). Other parasites reported in the 1996 study included the helminths Toxascaris leonina, Uncinaria stenocephala, Capillaria spp., Physaloptera spp., Paragonimus kellicotti, Alaria canis, Dipylidium caninum, Mesocestoides, and taeniid tapeworms, as well as the protozoans (Cysto)Isospora spp., Giardia spp., Hammondia heydorni, and Sarcocystis spp. Data were generated in these four studies using variations of flotation techniques including a passive zinc sulfate flotation test (or sometimes direct smears), zinc sulfate centrifugal flotation, or sucrose centrifugal flotation (Blagburn et al. 1996, 2014; Little et al. 2009; Mohamed et al. 2009).

It is well documented that centrifugal flotation techniques are more sensitive than passive flotation techniques, which can account for some prevalence variation in these studies (Zajac and King 2002; Zajac and Conboy 2012). Additionally, the choice of flotation solution and the specific gravity of that solution are important variables in parasite recovery (Zajac and King 2002; Zajac and Conboy 2012). Flotation techniques are not as efficient in detecting a variety of helminths with dense eggs and/or eggs that are susceptible to distortion in hypertonic flotation solutions, such as trematode, pseudophyllidean cestode, acanthocephalan, and some nematode eggs. The sedimentation technique is designed to recover dense eggs but is infrequently used for fecal diagnostic screening for companion animals (Zajac and Conboy 2012). Therefore, a combination of techniques including centrifugal flotation and sedimentation should increase the accuracy of parasite detection in population-based surveys or when evaluating individual animal samples for specific parasites based on differential diagnostic probabilities.

In this study, we estimated the prevalence of canine intestinal parasites, based on samples collected from geographically distinct animal shelters throughout Texas, using centrifugal fecal flotation and saline sedimentation techniques. In addition, we sought to identify risk factors for infection, as well as compare proportions of positive samples detected via fecal flotation vs. saline sedimentation for the most common parasites.

Materials and methods

Study design and sample collection

Sampling in this study was done in conjunction with two other research projects (Leahy et al. 2016; Hodo et al. 2019). This was a repeated cross-sectional study, in which dogs were sampled at 7 shelters across 7 Gould ecoregions of Texas (Gould et al. 1960). Each shelter was sampled 3 times, with approximately 30 dogs sampled at each shelter per visit, from May 2013 through December 2014. The animal shelters selected for inclusion were in the cities of Bryan/College Station, Dallas, Edinburg, El Paso, Fort Worth, Houston, and San Antonio.

Dogs 6 months of age and older with fresh feces available at the time of sampling were eligible for inclusion. When possible, we preferentially sampled dogs admitted to the shelter within the previous 48 h so that test results would reflect exposure in the dog’s original environment prior to entering the shelter. Deworming practices were variable; however, in most situations, dogs were dewormed on intake with pyrantel pamoate. Demographic data (age, sex, and dominant American Kennel Club [AKC] breed group) were estimated based on shelter records and/or the investigators’ assessments of the animal. Fecal samples were collected in the morning from the floor of kennels when dogs were housed separately. Dogs that did not have a fecal sample, or dogs co-housed, were walked on a leash in grassy areas to allow for defecation. Feces were scored for fecal consistency with a 1, 2, or 3 score. Feces with a score of 1 were formed, 2 were loose/semi-formed, and 3 were watery. Feces were stored in airtight plastic bags and kept on ice in coolers. Demographic data and fecal scores were compiled in a database along with fecal diagnostic test results. Our sampling protocol was approved by the Texas A&M University Institutional Animal Care and Use Committee, and informed consent was obtained from shelter directors.

Fecal diagnostic test methods

Centrifugal fecal flotation tests were performed on fecal samples within 1–3 days of collection using 2 g samples with commercial zinc sulfate flotation solution following previously described methods (Zajac and Conboy 2012). The coverslip was placed on the centrifuge tubes prior to a 5-min centrifugation, and the slides were examined using the 10 × objective. Any parasite eggs, cysts, or oocysts were recorded.

Fecal saline sedimentation tests were performed on fecal samples within 1–3 days of collection using 10 g samples with 1.2% NaCl solution following previously described methods (Zajac and Conboy 2012). Ten slides each, with 30 μl aliquots of sediment covered with a 25 × 25-mm coverslip, were evaluated using the 10 × objective. A higher NaCl concentration than what is standard for saline sedimentation procedures (0.85–0.9%) was used based on protocols used to recover Schistosoma sp. eggs from laboratory rodents to minimize hatching of the eggs during processing (Lewis et al. 2022). Because these fecal diagnostic tests were performed in conjunction with other research projects, not all samples had both fecal flotation and saline sedimentation performed.

Statistical analysis

Data were imported into a commercial statistical software program (SAS, version 9.4; SAS Institute Inc., Cary, NC) for variable coding and analysis. For each parasite, prevalence estimates and their 95% confidence intervals were calculated, stratified by fecal flotation vs. saline sedimentation. Descriptive analysis of the following variables of interest was performed: shelter location, age (< 1 year old or ≥ 1 year old), sex, duration of stay in shelter (in days), origin (stray or surrendered), and fecal sample consistency. Bivariable analysis using the chi-squared test was used to evaluate the independent relationship between potential risk factors and laboratory detection of each parasite. Separate multivariable logistic regression models for each parasite were used to identify risk factors for positive status. Initial selection of variables was based on the bivariable analysis screening (p < 0.25), and a backward elimination approach was used to establish the final models. The generalized estimating equations (GEE) method was used in these models to control for shelter as a random effect. In addition, McNemar’s chi-squared testing was used to compare proportions of positive samples detected via fecal flotation vs. saline sedimentation, as well as the combination of methods vs. each separately, for Ancylostoma caninum, Toxocara canis, and Trichuris vulpis. For all analyses, p values < 0.05 were considered statistically significant.

Results

Study sample data

Fecal samples were collected from 529 dogs for parasite evaluation. Fecal flotation was performed on 276 samples, whereas saline sedimentation was performed on 518 samples. Both methods were performed on a total of 265 samples.

Distribution of samples by shelter ranged from 57 (10.8%) to 85 (16.1%). Among all sampled dogs, 256 (48.4%) were female and 272 (51.4%) were male, with sex not recorded for 1 (0.2%) dog. A total of 451 (85.3%) were recorded as adults (≥ 1 year old) and 71 (13.4%) as puppies (< 1 year old), with age group not available for 7 (1.3%) dogs. The origin of 310 (58.6%) dogs was classified as stray, whereas 86 (16.3%) dogs were relinquished; origin was not available for 133 (25.1%) dogs. Duration of stay in the shelter prior to sample collection was 2 days or fewer for 215 (40.6%) dogs and longer than 2 days for 314 (59.4%) dogs.

Prevalence estimates

Parasite prevalence estimates are listed in Table 1. Overall, parasites identified by centrifugal fecal flotation (out of 276 dogs) were Ancylostoma caninum (26.4%), Toxocara canis (4.0%), Trichuris vulpis (12.0%), and Cystoisospora spp. (3.3%). Parasites identified by fecal saline sedimentation (out of 518 dogs) were A. caninum (20.7%), T. canis (3.9%), T. vulpis (14.1%), taeniid eggs (1.2%), Dipylidium caninum (4.1%), Alaria sp. (0.4%), Toxascaris leonina (0.2%), and Acanthocephalan eggs (0.4%).

Table 1.

Overall prevalence of intestinal parasites among dogs in 7 animal shelters across Texas, with samples tested via fecal flotation (276 dogs) and/or saline sedimentation (518 dogs)

Parasite	Fecal flotation, % (95% CI)	Saline sedimentation, % (95% CI)
Ancylostoma caninum	26.4% (21.3–32.1%)	20.7% (17.2–24.4%)
Toxocara canis	4.0% (2.0–7.0%)	3.9% (2.4–5.9%)
Trichuris vulpis	12.0% (8.4–16.4%)	14.1% (11.2–17.4%)
Taeniid	0%	1.2% (0.4–2.5%)
Dipylidium caninum	0%	4.1% (2.5–6.1%)
Alaria sp.	0%	0.4% (0.05–1.4%)
Toxascaris leonina	0%	0.2% (0.005–1.1%)
Acanthocephala	0%	0.4% (0.05–1.4%)
Cystoisospora spp.	3.3% (1.5–6.1%)	0%

Among samples tested using both methods, the detection of A. caninum, T. canis, and T. vulpis eggs was similar between fecal flotation and sedimentation techniques (Table 2). There was no significant difference in proportion of positive samples detected via fecal flotation vs. saline sedimentation for any of these three parasites. However, the proportion of A. caninum-positive samples detected by the combination of methods was significantly higher than the proportion detected by fecal flotation (p = 0.0009) or saline sedimentation (p = 0.0009) alone. The proportion of T. canis-positive samples detected by the combination of methods was significantly higher than the proportion detected by fecal flotation (p = 0.003) or saline sedimentation (p = 0.03) alone. Similarly, the proportion of T. vulpis-positive samples detected by the combination of methods was significantly higher than the proportion detected by fecal flotation (p = 0.0002) or saline sedimentation (p = 0.002) alone.

Table 2.

Comparison of centrifugal fecal flotation and fecal saline sedimentation detection of Ancylostoma caninum, Toxocara canis, and Trichuris vulpis among Texas shelter dogs for which both techniques were performed (N = 265)

Parasite	Fecal flotation, % (n)	Saline sedimentation, % (n)	Combined, % (n)
Ancylostoma caninum	25.7% (68)	25.7% (68)	29.8% (79)
Toxocara canis	3.8% (10)	5.3% (14)	7.2% (19)
Trichuris vulpis	12.1% (32)	13.6% (36)	17.4% (46)

Fecal score was collected for 89.0% (471/529) of samples. The median fecal score was 1, and the mean was 1.3.

Risk factor assessment

Dogs sampled within 2 days of admission were more likely to be shedding A. caninum than dogs sampled more than 2 days after being admitted (OR = 2.9, p < 0.0001), after controlling for shelter as a random effect. Dogs sampled within 2 days of admission were also more likely to be shedding T. canis than dogs sampled more than 2 days after being admitted (OR = 4.3, p < 0.0001). Dogs with abnormal fecal consistency (semi-formed or watery; fecal score 2 or 3) were more likely to be shedding T. vulpis than dogs with normal fecal consistency (OR = 1.8, p = 0.005). Adult dogs were more likely to be shedding T. vulpis than puppies (OR = 2.1, p = 0.03). Puppies (6 to 12 months of age) were more likely to be shedding taeniid eggs than adult dogs (OR = 3.3, p = 0.02). No significant multivariable associations were evident for the other parasites.

Discussion

Most shelter dog intestinal parasite prevalence studies have utilized only fecal flotation methods. The addition of fecal sedimentation not only added to the number of parasite species detected in this study, but it also increased the number of dogs diagnosed with the common intestinal parasites that are primarily detected via fecal flotation methods. Fecal flotation methods have the benefit of concentrating parasite eggs that are less dense than the flotation solutions used and allow for examination of the slide with minimal debris in the sample; however, these methods can distort parasite eggs/cysts that are susceptible to collapse in hypertonic solutions. Fecal sedimentation techniques allow for the detection of more dense parasite eggs but are not as efficient at concentrating eggs, and eggs must be discerned among a large amount of debris from the sample. It was unexpected by the authors that fecal flotation and sedimentation would yield similar proportions of positive samples detected. This could have been impacted by higher parasite burdens and fecal egg numbers, leading to the increased ability to detect a given egg on fecal sedimentation. It should be noted that the quantity of fecal material processed in the fecal sedimentation is larger than the amount used in fecal flotations, which impacts the total number of eggs that are available for detection in sedimentation tests. Our results indicate that combining fecal flotation and saline sedimentation will enhance detection of Ancylostoma caninum, Toxocara canis, and Trichuris vulpis eggs. However, we cannot rule out that non-uniform parasite egg distribution in the fecal samples could have contributed to the detection of parasites with one test and the failure to detect in the other. It is possible that performing fecal flotations in duplicate could have resulted in a similar enhancement of detection of these parasites due to non-homogeneous distribution of parasite eggs in the samples.

The prevalence estimates of A. caninum (26.4%) and T. vulpis (12.0%) by fecal flotation alone were similar to those of previous canine shelter studies (Blagburn et al. 1996, 2014). The prevalence of T. canis (4.0%) was lower than in those studies (14.5% and 12.5%, respectively) and may have been influenced by the fact that most dogs sampled in this study were adults (85.3%). The hookworm, whipworm, and roundworm prevalence estimates in these shelter dogs were higher overall than the prevalence reported by Companion Animal Parasite Council in 2014 for Texas, which were 3.18%, 0.65%, and 1.19% respectively (CAPC 2014), demonstrating that shelter dogs carry a higher burden of intestinal parasites compared to pet dogs.

In this study, we detected additional parasites (Taeniidae, Dipylidium caninum, Acanthocephalan, Alaria sp., and Toxascaris leonina) that would not have been detected if we had only performed the ZnSO₄ centrifugation fecal flotation procedure. Taeniid, Dipylidium caninum, and Acanthocephalan eggs can be detected on fecal flotation, although the higher specific gravity of these eggs makes recovery with flotation not as efficient as with less dense eggs. Alaria sp. eggs can also be detected on fecal flotation, but they can be distorted and/or not recovered due to the collapse of these eggs in hypertonic solutions (Nagamori et al. 2020). In a national shelter intestinal parasite prevalence study that used sugar centrifugation only as the method of detection, other helminth eggs detected besides hookworm, roundworm, and whipworm were Capillaria sp. (0.39%), Physaloptera sp. (0.05%), Paragonimus kellicotti (0.02%), Alaria canis (0.08%), D. caninum (0.09%), Mesocestoides sp. (0.03%), and Taeniidae (0.6%) (Blagburn et al. 1996). The prevalence estimates for D. caninum and Taeniidae in that study were likely underestimated. In one canine shelter study with necropsy gold standard comparison, sugar centrifugal flotation detected only 3/48 dogs with D. caninum and 4/7 dogs with Taenia sp. worms present in the intestinal tract (Adolph et al. 2017).

Commercial sugar flotation solution has a higher specific gravity (1.27) compared to commercial ZnSO₄ solution (1.18), and sugar centrifugation was demonstrated to be significantly better at recovering T. vulpis eggs, the densest of the common helminth eggs, compared to ZnSO₄ centrifugation (Dryden et al. 2020). It is possible that if we had used sugar centrifugation, we could have also detected some of the parasites that were only detected by sedimentation. In addition, in our study, coverslips were immediately removed from the centrifugation tubes and placed on a slide for reading. It has been demonstrated that for centrifugation, egg recovery increases when the coverslip remains on the centrifugation tube for 10 min after spinning (Dryden et al. 2005). Had we allowed more time for the coverslips to remain on the centrifugation tubes, it is possible that flotation could have detected more parasite eggs in this study. Additionally, the fecal flotation procedure was performed by students who were trained on the technique for the purpose of this study, so it is possible that they were not as accurate in detecting eggs, cysts, or oocysts as a more experienced person would have been. This could explain why there were no Giardia sp. or Sarcocystis sp. identified, as these have been detected in other canine intestinal parasite surveys (Blagburn et al. 1996; Little et al. 2009).

While this study demonstrated the detection of more infected dogs and additional parasites by combining fecal flotation and sedimentation, there are other tests that could be performed to further increase the types of parasites detected in shelter studies. Canine lungworms including Aelurostrongylus abstrusus, Crenosoma vulpis, and Angyostrongylus vasorum are primarily detected by the Baermann technique. Fecal smears can detect trophozoites or cysts of trichomonads. Stained fecal smears can also detect Cryptosporidium spp. and trichomonads. Coproantigen tests have been developed to detect T. canis, A. caninum, and T. vulpis even prior to egg shedding (Sweet et al. 2021). Extraction of DNA and the performance of PCR using a variety of platforms can also be performed when targeting specific genera and/or species of parasites.

The original purpose of performing the fecal sedimentation procedure in this study was to estimate the prevalence of the trematode parasite Heterobilharzia americana in Texas shelter animals. Eggs were not detected in this population of shelter dogs; however, this parasite has been detected in hundreds of pet dogs in the state of Texas (Rodriguez et al. 2014). This parasite was also detected in one dog via fecal sedimentation in a dog park fecal survey in Oklahoma, USA (Duncan et al. 2020). It is possible that this parasite was missed if infected dogs had low egg shedding, but overall, we can conclude that H. americana was uncommon in this population of shelter dogs in Texas. Due to the time intensity of performing fecal sedimentations on a large number of samples (in this study, approximately 4–6 h for sedimentation and 20 min of reading slides per sample), fecal PCR may be a preferred method of estimating H. americana prevalence in large dog populations.

Detailed deworming history was not available for each individual dog sampled; however, pyrantel pamoate given at or near the time of intake was the most common shelter deworming protocol. This could explain why dogs had a lower risk of A. caninum and T. canis infections if they were sampled after 48 h of shelter admittance. In contrast, it is not surprising that there was no difference in T. vulpis infection between dogs admitted within 48 h or greater than 48 h, as pyrantel pamoate is not labeled as an effective treatment for this nematode. Interestingly, the presence of T. vulpis correlated with abnormal fecal scores, which suggests that shelter dogs with abnormal feces should receive broader anthelmintic treatment to address T. vulpis infection, especially when fecal diagnostic capabilities are limited within the shelter. Although not demonstrated to be zoonotic, T. vulpis eggs are persistent in the environment, and these infections may cause chronic gastrointestinal disease and in severe cases may lead to death (Traversa 2011). Because of the relatively high prevalence of this parasite in these shelter animals, shelters should consider increasing the spectrum of their deworming protocols if financially feasible.

Regardless of the shelter deworming procedures, veterinarians examining newly adopted shelter animals should perform fecal diagnostic testing and administer broad spectrum anthelmintic treatment to these patients to address untreated and/or undiagnosed whipworm and tapeworm infections. Pet owners should also be educated on fecal sanitation in their home environment to prevent contamination with parasite eggs and reinfection or infection of other dogs. Contamination of the home environment with T. canis, A. caninum, and D. caninum also pose a zoonotic risk. Furthermore, although A. caninum infections are typically well controlled by anthelmintic treatment, cases of recurrent/persistent canine hookworm infections are increasingly reported. Recent experimental hookworm infections with parasites from dogs with persistent infections across the southeastern USA showed the presence of multiple drug resistance to benzimidazoles, macrocyclic lactones, and pyrantel in A. caninum, suggesting an emerging resistance problem (Jimenez Castro et al. 2019).

Knowledge of the endoparasite burden in shelter dogs is important not only for enhanced veterinary understanding but also for advancing public health. For example, eggs of T. canis and Toxocara cati were identified in soil from 29.6% of New York City area playgrounds (Tyungu et al. 2020). Human exposure to these roundworms disproportionately impacts poor and minority populations, causing cognitive and developmental delays, lung dysfunction, and asthma, and is potentially linked to a substantial yet hidden burden of mental illness in the United States (Hotez 2014). Given our findings of Toxocara eggs in 7.2% of the Texas shelter dog samples on which both flotation and sedimentation tests were performed, the human medical community should be aware of potential exposures in areas where these dogs frequent. A One Health approach is increasingly critical to maximize the impact of veterinary surveillance data.

Conclusions

Combining fecal sedimentation with fecal flotation in this shelter dog intestinal parasite study increased the overall number of dogs detected with intestinal parasites. Fecal sedimentation also detected tapeworm, trematode, and acanthocephalan parasites that would not have been detected if only fecal flotation had been performed in this study. Texas shelter dogs carry a high burden of intestinal parasites, including those of zoonotic importance. In performing future intestinal parasite survey studies, incorporating fecal sedimentation techniques and other tests in addition to standard fecal flotation methods may help increase the accuracy of intestinal parasite prevalence estimates. Because of the wide range of intestinal parasites present in shelter dogs, broad spectrum deworming protocols should be considered if costs are not prohibitive. Veterinarians should ensure that proper fecal examination and broad-spectrum antiparasitic treatment occur following adoption. Because shelter dogs are at higher risk of exposure to a broad variety of parasites, incorporating fecal sedimentation and other diagnostic methods that can increase sensitivity of parasite detection and/or increase the types of parasites detected should be considered, in addition to sugar centrifugation flotation.

Data availability

The dataset generated and analyzed during the current study is available from the corresponding author upon reasonable request.