Anthelmintic resistance in equine gastrointestinal nematodes is a threat to equine health and welfare. Detailed knowledge of anthelmintic use and parasite control methods is a prerequisite to identification of potential risk factors for resistance.
To identify parasite control practices employed by equine owners in Scotland and investigate management factors associated with anthelmintic resistance.
Questionnaire study of equine parasite control in Scotland.
Questionnaires were available electronically, distributed at a conference and mailed to clients. Key areas explored included general background, grazing management, anthelmintic treatment practices and use of diagnostic tests.
A total of 193 responses detailing information on parasite control programmes of 993 equids were analysed. Moxidectin (MOX) and ivermectin or related combination products were the most commonly administered anthelmintics in the preceding 12 months. Treatments licensed for use against cyathostomin encysted larvae and tapeworms were administered by 80% and 90% of respondents, respectively. This was often achieved through indiscriminate use of MOX and MOX-praziquantel products. Faecal egg count (FEC) analysis had been performed on 62% of yards and regular use of FECs reduced annual anthelmintic treatment frequency. Veterinarians had the greatest influence on control practices. While 40% of respondents believed that they practised targeted dosing, this was not associated with delaying treatment beyond the egg reappearance period of the anthelmintic used.
Responses indicated increasing veterinary involvement and use of FECs. The majority of respondents administered anthelmintics licensed against cyathostomin encysted larvae and tapeworms. However, responses suggested that owners did not understand the definition of ‘targeted’ dosing regimens.
The high frequency of MOX use represents a potential risk factor for macrocyclic lactone resistance. As veterinarians were the most influential factor in anthelmintic choice, awareness of macrocyclic lactone resistance and potential risk factors for its development and spread should be incorporated into client advice.
All grazing equids are at risk of gastrointestinal parasitism, representing an important threat to their health and welfare . Cyathostomins are the most important gastrointestinal helminths, considering their high prevalence, potential pathogenicity and anthelmintic resistance . Presently, 3 classes of anthelmintic are available for their control; benzimidazoles (BZD), tetrahydropyrimidines (THP) and macrocyclic lactones (ML). However, widespread cyathostomin resistance to BZD and THP is present [3, 4] and reduced efficacy of ivermectin (IVM) [5, 6] and moxidectin (MOX)  has been recently reported. No additional equine anthelmintic classes are expected in the near future; therefore it is essential that efficacy of currently effective products is preserved. There is a requirement for more sustainable, targeted control practices, based on sound epidemiological principles, which identify individuals requiring anthelmintic treatment, test the sensitivity of anthelmintics administered and rely more on pasture management of infection load .
Three main helminth control protocols have been described; interval, strategic and targeted dosing . Interval dosing involves administration of anthelmintics at set times determined by the egg reappearance period (ERP) of the anthelmintic used. This regimen, proposed in the 1960s, has substantially reduced parasite-associated disease in equids, but has also accelerated anthelmintic resistance . Strategic dosing involves administration of anthelmintics at particular times of the year based on parasite life cycles . This does not account for variation in parasite burden and nematode egg excretion, which can differ markedly between individuals , or for abnormal weather patterns influencing pasture infection load . Targeted dosing recognises that parasites are likely to be aggregated at host level, with only a small percentage of the equine population carrying a high burden, and most individuals having low or zero faecal egg counts (FECs) . Equids with a moderate/high FEC are treated at appropriate times as determined by parasite life cycle and epidemiology . Remaining individuals are left untreated, leaving a proportion of the in-host parasite population in refugia i.e. unexposed to anthelmintic , reducing unnecessary anthelmintic administration . Targeted dosing is currently considered the best control programme for equine gastrointestinal helminths . It has been recommended that targeted dosing should involve the use of FECs through the grazing season and annual dosing against cyathostomin encysted larvae (EL) and tapeworms . Anthelmintics currently licensed for use against cyathostomin EL in the UK are MOX and 5-day fenbendazole (5d-FBZ). Anthelmintics licensed to treat tapeworms are praziquantel (PRAZ) and pyrantel (PYR) at twice the nematocidal dose.
Targeted equine parasite control programmes have been advocated for many years  but are not widely implemented. Questionnaire studies from the last 2 decades indicate that most equine premises over-use anthelmintics through continued interval dosing [16, 17]. Further investigation of the use of anthelmintics and helminth control practices is required to determine risk factors associated with the development and spread of anthelmintic resistance and to identify barriers preventing the implementation of more rational control methods.
Materials and methods
Questionnaires were distributed in Scotland during 2010 by 3 methods; a postal survey of clients of the Edinburgh University Equine Practice, dissemination to delegates at the British Horse Society (BHS) Scottish Welfare Conference and a web-based survey of Scottish BHS members.
The questionnaire comprised 47 questions: 8 open- and 39 close-ended multiple-choice questions with the opportunity to include additional comments at the end (questionnaire available as Appendix S1). Several key areas were explored: general information, grazing and parasite control practices, and the use of diagnostic tests. Where applicable, questions were subdivided into further age-based categories; foals (<1 year), youngsters (1–3 years) and adults (>3 years), with an additional category for donkeys. In answer options, anthelmintics were detailed by chemical ingredient rather than by class.
Microsoft Excel was used for data analysis and descriptive statistics. In calculating percentages, the denominator excluded any missing responses and results were rounded to the nearest integer.
Respondents were asked whether they performed specific planned dosing of their equids against cyathostomin EL and tapeworms and which anthelmintics were used. This ‘perceived’ specific planned dosing prevalence was compared with the ‘actual’ dosing prevalence derived from analysis of the anthelmintics each respondent listed having used in the preceding 12 months. Two-by-two contingency tables were used to detect any significant difference in ‘perceived’ vs. ‘actual’ dosing against cyathostomin EL and tapeworms, employing Chi-squared and 2-tailed Fisher's exact tests.
The relationship between management factors and treatment frequency in adult equids was explored using multiple linear regression. The number of annual treatments in adult equids was the response variable, and questionnaire responses were included as fixed factors for categorical variables (for example, whether or not faeces were removed from pasture) on the basis that all possible responses were included, and random factors for scale and ordinal variables (for example, number of equids on yard; frequency of faecal removal). Respondent source was included as a dichotomous variable because nominal data in more than 2 categories are not suitable for multiple regression; hence ‘university practice client’ or ‘other’ (i.e. welfare conference delegates and web respondents). For pasture rotation, pasture resting and faecal removal, responses were coded into ordinal categories of decreasing frequency. For frequency of FECs, responses were coded in order of increasing frequency. Additional factors were included to reflect stocking density (i.e. total number of equids divided by total grazing area in hectares) and age profile (proportion of young equids and foals). Only continuous scale variables, ordinal variables with 5 or more levels, and binomial categorical variables were included as factors. All factors were included in the initial model and the least significant removed in turn until either only significant factors remained (at α = 0.05), or the removal of further factors led to a previously significant P value rising above 0.05. In the final model, only significant factors were reported. Too few young equids, foals and donkeys were included in the study to justify a similar analysis within these separate categories.
The total number of treatments given provides a limited perspective on treatment intensity, since the residual effect of different anthelmintics varies. Less frequent use of a longer-acting anthelmintic such as MOX could therefore exert strong selection pressure on parasite populations even while being used much less frequently than an anthelmintic with little or no residual effect. To address this, the expected duration of ‘effect’ (as measured by accepted standard strongyle ERP) of the most commonly used anthelmintic was subtracted from the reported treatment frequency. If a positive value remained, this was taken to indicate increased provision of a ‘treatment-free period’ (TFP). Expected duration of effect was taken to be 13 weeks for MOX, 8 weeks for IVM, and 6 weeks for FBZ and PYR as per manufacturers' recommendations of ERP . For this analysis, treatment frequency was interpolated from the ranges offered in the questionnaire; for example 7–9 weeks became 8 weeks. The relationship between the presence of a TFP and management factors as described above was assessed using logistic regression. Respondent source was included as a nominal variable with 3 categories: University practice clients, welfare conference delegates and web respondents, followed by 2 categories: university practice clients and ‘other’ (i.e. welfare conference delegates and web respondents). The final model retained significant factors only. Regression analyses were conducted in PASW 18.0a.
Relationships between management factors were assessed using Spearman rank correlation. Bonferroni correction was applied to adjust the critical P value to compensate for multiple tests, by dividing α = 0.05 by the number of comparisons made. Only ordinal and scale variables were included in this analysis.
Influence of veterinarians on parasite control programmes was compared between university practice clients and ‘other’ using a Chi-squared test. Significance was indicated by P<0.05.
Questionnaire response rate and demographic data
Of the 213 responses received, 20 web-based responses in which <50% of questions were answered were excluded, providing a usable response rate of 91%. Therefore, a total of 193 responses were analysed: 75 (39%) from the postal survey of clients of the Edinburgh University Equine Practice, 18 (9%) from conference delegates and 100 (52%) from the web-based survey. Missing data ranged from 0 to 43% of total respondents per question, with a median of 7%. Internet publication of the questionnaire precluded calculation of a true response rate.
Respondents directly owned/loaned 993 equids (median 2, range 1–130), representing approximately 1.2% of the Scottish equine population of 80,000 (personal communication, Helene Mauchlen, Scottish BHS). Of these equids, 91% were mature (>3 years), 6% youngsters (1–3 years), 2% foals and 1% donkeys. They detailed control practices for 3069 equids on the yards surveyed (mean 16/yard, median 8, range 1–130). Responses were mostly from private (47%) and livery (38%) yards, with smaller numbers from riding schools and multipurpose yards. All equids had access to grazing.
Anthelmintic treatments and FEC analysis
All data refer to anthelmintic treatments in the preceding 12 months. Moxidectin and IVM, or their combination products with PRAZ (i.e. MOX-PRAZ, IVM-PRAZ) were the most commonly used anthelmintics, representing 43% and 33% of the total products used, and used on 92% and 71% of yards, respectively. PYR and FBZ comprised 10% and 8% of the total anthelmintics used, on 21% and 17% of yards, respectively (Fig 1). Although only 57% of respondents perceived that they treated specifically for cyathostomin EL, 80% actually reported using a larvicidal anthelmintic (Fig 2a); 90% of whom used a MOX-based product and 10% 5-day FBZ. Similarly, despite only 78% of respondents perceiving that they were administering anthelmintics specifically against tapeworms, 90% reported using an anticestode anthelmintic (Fig 2b); single (12%) or combination (72%) PRAZ products were used more frequently than PYR (16%). Forty-four percent of respondents stated they used strategic, 40% targeted and 15% interval dosing (Fig 3a), following the definitions detailed in the questionnaire. Most anthelmintics were administered every 13–15 weeks (22%) or every 4–6 months (20%), with fewer equids being treated more frequently and 19% specifically stating they dosed ‘based on FEC’ rather than specific dosing intervals (Fig 3b). For anthelmintic dose calculations, objective weight estimation was performed by 73% respondents; 70% using a weigh tape and 3% weigh scales. The remaining respondents estimated weight visually (15%), administered one tube per equid (11%) or were unaware of how the dose was determined (1%).
Faecal egg count analysis had been performed on 62% of yards surveyed; every 2–3 months on 33% of those yards and every 4–6 months on 29% (Fig 4). The multiple regression model captured 32.3% of the variance in annual number of treatments given to adults (Table 1). Fewer treatments were given to mature horses on yards with a higher proportion of young equids, and on yards where treatment was stated to be ‘targeted’ and/or ‘based on FEC’. Yards with equids at grass for a greater part of the day in spring treated their equids less overall, whereas equids that grazed for a greater part of the day in winter were treated more frequently. Increasingly frequent FEC analysis, from never to every 2–3 months or more, was associated with fewer annual treatments overall.
Table 1. The relationship between management factors and the annual number of treatments given to adult horses, assessed using multiple linear regressiona
aOnly significant factors are included. ANOVA F11,131 = 6.670, P<0.001. Model constant is not shown. Targeted treatment and treatment based on faecal egg counts (FECs) were dichotomous variables, and frequency of FEC an ordinal variable with 0 = FEC not conducted, 1 = conducted infrequently, 2 = every 4–6 months, 3 = every 2–3 months or less. ‘Targeted treatment’ represented respondents who stated they practised targeted treatment, rather than interval or strategic treatment, following the definitions provided in the questionnaire. ‘Treatment based on FEC’ represented respondents who specifically answered in a separate question that they were using FEC to determine whether or not to administer an anthelmintic, rather than using set treatment intervals. Respondent source compared responses from university practice clients and ‘other’ (welfare conference delegates and web-respondents).
Proportion of young horses
Spring grazing (h/day)
Winter grazing (h/day)
Treatment based on FEC
Frequency of FEC
Management factors and the treatment-free period
Treatment-free period was estimated for 137 yards. Of these, 51 (37%) administered their most commonly used anthelmintic less frequently than its ERP, thereby providing a TFP. The median duration of this TFP was 6 weeks (10th to 90th percentiles, 6–38). Logistic regression revealed yards that harrowed more frequently were more likely to delay treatment beyond the expected period of strongyle egg suppression (Table 2). This delay was also 7.5 times more likely to be found on yards that had used FEC frequently and nearly 10 times less likely on those that stated that they used ‘targeted treatment’.
Table 2. Logistic regression of management factors on the presence of a treatment-free period (see text) between routine treatmentsa
aHosmer and Lemeshow Chi-squared 7.486, 8 degrees of freedom, P = 0.485. Stocking density (not shown) was retained as a supporting factor (B = 0.045, P = 0.056), whose removal reduced the P value of other factors below 0.05. OR = odds ratio; CI = confidence interval. ‘Targeted treatment’ represented respondents who stated they practised targeted treatment, rather than interval or strategic treatment, following the definitions provided in the questionnaire. ‘Treatment based on FEC’ represented respondents who specifically answered in a separate question that they were using faecal egg counts (FECs) to determine whether or not to administer an anthelmintic, rather than using set treatment intervals. Respondent source compared responses from university practice clients and ‘other’ (welfare conference delegates and web-respondents).
Treatment based on FEC
Quarantine and pasture management
Fifty-seven percent of respondents knew whether or not their premises had a quarantine policy, and quarantined new arrivals for a median of 10 days (range 1–42). Seventy-six percent of respondents removed faeces from pasture, with 39% removing faeces at least twice weekly. Of 137 people who detailed the method of removal, 85% removed faeces manually and 11% by machine.
Correlations between management factors
Thirteen factors were included in correlation analysis, giving a Bonferroni-adjusted critical P value of 0.004. Yards with higher stocking densities had more equids in total (Spearman ρ = 0.39, P<0.001) and proportionately more young equids (ρ = 0.27, P<0.001). Yards with more equids in total also had a younger age profile (ρ = 0.43, P<0.001). The time for which pastures were rested was negatively correlated with the total number of equids present (ρ = -0.22, P = 0.003). The frequency of pasture rotation, harrowing and faecal removal were not significantly associated with any of the other management factors tested. The hours/day that equids grazed in spring, summer, autumn and winter were all positively intercorrelated (ρ = 0.69–0.86, P<0.001).
Clinical signs associated with parasitism
In the preceding 12 months, 8% of respondents observed clinical signs consistent with ‘worm-related disease’ including ‘worms in faeces’ (36%) and ‘weight loss’ (32%). A veterinary diagnosis was made in 63% of these cases, using ‘other’ methods such as history taking or clinical examination (50%), FEC (30%) or tapeworm ELISA (20%).
Influences on parasite control programmes
Veterinarians and yard owners were the main influence on parasite control practices according to 40% and 32% of respondents, respectively. There was no significant difference in the stated influence of veterinarians on parasite control between respondent sources (P = 0.89).
Perceptions of anthelmintic resistance
The majority (86%) of respondents stated that they were unaware of anthelmintic resistance on their yards. The median response value for concern regarding resistance on a scale of 1–10 (10 being most concerned) was 6. Sixty-one percent claimed to be aware of what a faecal egg count reduction test (FECRT) was, and of these, 16% reported that a FECRT had been performed at their premises.
The threat of widespread anthelmintic resistance in equine helminths necessitates a change in approach to anthelmintic use, to preserve anthelmintic efficacy and minimise parasite-associated disease. Awareness of control programmes employed and owners' knowledge, and compliance is essential to accurately direct future advice. Several questionnaire studies conducted in the UK and EU were performed prior to registration of MOX, when IVM was the predominant anthelmintic used [18, 19]. MOX is now an important tool in control programmes due to its licensed efficacy against cyathostomin EL, especially on premises with a high prevalence of resistance to BZD .
This is the first questionnaire study of anthelmintic use in horses conducted solely in Scotland, and the number of respondents exceeds that reported in previous similar studies in other parts of the UK [16, 17, 19]. The study population comprised a similar demographic to the general managed equid population in Scotland, where 50% of equids are kept on private premises and the remainder on rented premises (personal communication, Helene Mauchlen, Scottish BHS). It must be acknowledged, however, that an inherent bias exists in the results of any voluntary questionnaire study, as respondents may be conscientious or more interested in the subject matter than the general population.
The results revealed a predominance of ML use, with MOX and IVM, or their combination products, representing 43% and 33% of the total products used in the preceding 12 months, respectively. These results are similar to those of , where MOX comprised 39% of total anthelmintics administered, verifying a shift in preference from IVM to MOX in the UK. This is likely to be due to the larvicidal activity of MOX, its longer standard ERP and strong marketing initiatives. With widespread resistance to BZD and THP [3, 4], and reports indicating reduced efficacy of ML against cyathostomins [6, 7], indiscriminate use of ML could potentially culminate in resistance to all available anthelmintics. Furthermore, considering the continued efficacy of PYR in Scotland , sole reliance on ML is not currently indicated.
The proportion of respondents treating cyathostomin EL and tapeworms was 80% and 90% respectively. These exceed the 35% and 75%, and the 64% and 74% treatments described in earlier studies of Thoroughbred trainers and eventers, respectively [16, 17]. However, these studies only considered 5d-FBZ and double dose PYR as treatment options for cyathostomin EL and tapeworms, respectively. Moxidectin and PRAZ were preferentially used by yards in this study, probably reflecting reports of cyathostomin resistance to FBZ, including the 5-day course [20, 23]. The statistically significant difference observed here, between ‘perceived’ and ‘actual’ treatments administered for these stages or species of parasite, indicates that some owners remain unaware of the spectrum of activity of anthelmintics used, and is likely to reflect high year-round use of MOX and MOX-PRAZ products. Indiscriminate use of MOX provides a risk factor for ML resistance. Therefore, it could be argued that MOX administration should be reserved for specific treatment against cyathostomin EL and highlights a requirement for continued client education. The high frequency of MOX use is likely to be responsible for the reduced frequency of treatments observed, with most anthelmintics being administered every 13–15 weeks or 4–6 months. This contrasts with previous studies of Thoroughbred racehorses and pleasure horses where the majority of equids received interval treatments every 4–6 weeks [16, 24].
The responses indicate increased implementation of strategic (44% respondents) or targeted control programmes (40% respondents). Variation in the apparent influence of a targeted approach on treatment practices suggests that respondents were unclear of the definition of this term; for example, while yards claiming to use a targeted approach administered fewer treatments to mature horses, this did not increase the likelihood of delaying treatment beyond the expected duration of effect of the anthelmintic used. Targeted treatment meant different things to different respondents, highlighting the need for continued knowledge transfer, with a consistent definition of targeted treatment within the veterinary literature and lay press, notably encouraging the use of FEC to guide anthelmintic use.
Use of FECs to target treatments has been shown to reduce anthelmintic administration by up to 75% . Faecal egg counts are fairly inexpensive and easy to perform. The relatively high level of use of FECs in this study is encouraging, particularly compared with previous observations where FEC analysis was mostly employed to investigate ‘problem’ cases [16, 24]. Indeed, 64% yards performed regular (i.e. > q. 4–6 months) FECs, with only 1% of FECs performed to investigate ‘problem’ cases. Those respondents conducting FECs more frequently administered fewer anthelmintics to adult equids/year on average. Similarly, those dosing on the basis of FEC rather than at a set dosing interval were 7 times more likely to provide a TFP. However, it is important that the diagnostic limitations of FECs including their methodology and interpretation be appreciated to best integrate their use into control protocols .
Despite the presence of anthelmintic resistant nematodes for several decades , there has been concern regarding poor dissemination of this information among equine veterinarians and owners . Here, 86% of respondents stated that they were unaware of the presence of resistance on their yard, mirroring the finding of O'Meara and Mulcahy  that only 16% of respondents were aware of FECRT. A recent study found that owners in the UK represented a receptive audience for efficacy testing, with 46% of respondents willing to pay at least £2/month to test for resistance . Thus, indicating that the FECRT should be promoted further as a means of monitoring anthelmintic efficacy and be incorporated into control protocols.
To limit the spread of anthelmintic resistance, it is essential to prevent the introduction of resistant parasites onto premises. Administration of a larvicidal and cestocidal anthelmintic to all new arrivals prior to pasture turnout should be encouraged. Anthelmintic treatment of ‘new arrivals’ was performed by 89% of yards studied, although the anthelmintic administered was often unknown. Improved pasture management is required to underpin successful control programmes. Removal of faeces from pasture reduces the number of free-living parasitic stages in the environment, thereby reducing the need for anthelmintic treatment . Forty-five percent of respondents removed faeces from pasture at least once weekly, with 39% removing them at least twice weekly. However, frequency of faecal removal showed no significant correlation with anthelmintic treatment frequency.
There were few correlations among different management factors, indicating that future yard-specific recommendations are required rather than a single uniform approach to parasite control on all premises. Yards with more equids tended to have higher stocking densities, more young animals and less scope to rest pastures. This presents a challenge to integrated management with reduced reliance on anthelmintics, although the economic benefits from reduced use would possibly be higher on such premises. Worryingly, yards with higher proportions of young stock gave fewer treatments to co-grazing adults, despite likely higher pasture infectivity on such premises due to higher infection intensities often present in young stock . Yards that grazed equids for more of the day tended to do so in all seasons. Spring grazing was associated with reduced overall treatment, and winter grazing with more. This is probably due to high use of larvicidal treatments with MOX in the winter months and its long ERP leading to an associated reduction in anthelmintic treatments in the subsequent spring.
Under-dosing is an avoidable risk factor for resistance [28, 29]. Previous studies showed objective weight estimation was under-used, with 87–94% owners relying on visual estimation of weight alone [30, 31]. In this study, 73% of respondents used some form of objective weight estimation. This aligns with a recent UK study where 88% and 10% respondents used a weigh tape or weighbridge, respectively .
Encouragingly, this study indicates that equid owners are aware of the importance of helminth control as previously described . In general, the respondents appeared to have successfully prevented parasite-associated disease, with potential cases of clinical disease evident on only 8% of yards. This contrasts with previous studies where the incidence of nematode-associated disease ranged from 18 to 40% [16, 19], and a recent study of UK stud farms where a incidence of 26% was observed . However, in many cases of suspected parasite-related disease, a definitive diagnosis was not reached, reflecting a paucity of appropriate tests to accurately evaluate nematode burden, and consequent inaccuracies in the determination of true prevalence. This emphasises the requirement for improved diagnostic tests to aid clinical evaluation and decision-making.
The high level of veterinary involvement in helminth control found in this study is encouraging and possibly contributed to the increasing implementation of FECs. Previously, such influence was only reported on a minority of racing yards in the UK [16, 33]. Interestingly, however, clients of the university practice did not report being significantly more influenced by veterinarians than respondents from the other sources and in fact used anthelmintics more frequently and were less likely to provide a TFP. This particularly highlights the need for veterinary practices to provide more integrated parasite control advice to their clients. In a concomitant study of UK stud farms using a similar questionnaire  75% of respondents took veterinary advice; however, the continued practice of interval dosing was common . This may reflect the differing age distributions of the 2 study groups, with stud farms having a greater proportion of young stock and increased movement on and off premises. The high monetary value of Thoroughbreds is also likely to lead to an increased fear of parasite-associated disease.
In summary, this study revealed implementation of more rational parasite control protocols on Scottish yards, with increased use of FECs and veterinary involvement. The frequent use of MOX represents a potential risk factor for ML resistance. The apparent misunderstanding of the definition of the different control protocols requires clarification and implementation of FECRT should be encouraged to monitor anthelmintic efficacy.
Authors' declaration of interests
No competing interests have been declared.
Ethical animal research
No animals were included in this study.
Sources of funding
This work was supported by funding from the Horse Trust, The Horserace Betting Levy Board (HBLB) and Elise Pilkington Trust.
C.S. is a Horse Trust Clinical Scholar in Equine Medicine. The authors would like to thank all the individuals that participated in this questionnaire, and the Scottish BHS for their help with questionnaire distribution.
All authors contributed to study design. Data collection was performed by Claire Stratford. Data analysis was performed by Claire Stratford, Hannah Lester and Eric Morgan, and interpretation by Claire Stratford, Hannah Lester and Jacqui Matthews. All authors contributed to the writing of the manuscript.