Search

Abstract Background Fasciolosis (Fasciola hepatica) and paramphistomosis (Calicophoron daubneyi) are two important infections of livestock. Calicophoron daubneyi is the predominant Paramphistomidae species in Europe, and its prevalence has increased in the last 10–15 years. In Italy, evidence suggests that the prevalence of F. hepatica in ruminants is low in the southern part, but C. daubneyi has been recently reported at high prevalence in the same area. Given the importance of reliable tools for liver and rumen fluke diagnosis in ruminants, this study evaluated the diagnostic performance of the Mini-FLOTAC (MF), Flukefinder® (FF) and sedimentation (SED) techniques to detect and quantify F. hepatica and C. daubneyi eggs using spiked and naturally infected cattle faecal samples. Methods Briefly, negative bovine faecal samples were artificially spiked with either F. hepatica or C. daubneyi eggs to achieve different egg count levels: 10, 50 and 100 eggs per gram (EPG) of faeces. Moreover, ten naturally infected cattle farms from southern Italy with either F. hepatica and/or C. daubneyi were selected. For each farm, the samples were analysed individually only with MF technique and as pools using MF, FF and SED techniques. Bayesian latent class analysis (LCA) was used to estimate sensitivity and accuracy of the predicted intensity of infection as well as the infection rate in the naturally infected farms. Results The outcome of this study showed that the highest number of eggs (F. hepatica and C. daubneyi) recovered was obtained with MF, followed by FF and SED in spiked infected samples at 50 and 100 EPG, while at lower infection levels of 10 EPG, FF gave the best results. Moreover, the sensitivity for all the techniques included in the study was estimated at > 90% at infection levels > 20 EPG for both F. hepatica and C. daubneyi eggs. However, MF was the most accurate of the three techniques evaluated to estimate fluke infection intensity. Nevertheless, all three techniques can potentially estimate infection rate at farm level accurately. Conclusions Optimization and standardization of techniques are needed to improve the FEC of fluke eggs. Graphical Abstract

Abstract Background Several studies in both experimentally and naturally infected dogs have reported the adulticide effect of a combination of macrocyclic lactones and doxycycline against Dirofilaria immitis, showing that these protocols can be used as an alternative to melarsomine. The present study evaluated the efficacy of oral, topical and extended-release injectable formulations of moxidectin when combined with doxycycline in dogs naturally infected with D. immitis from a shelter located in southern Italy. Methods Thirty dogs with naturally acquired D. immitis infection were divided in three groups (G) and treated with oral moxidectin (G1) once a month for 9 consecutive months, topical moxidectin (G2) once a month for 9 consecutive months or extended release moxidectin injectable (G3) at enrolment and again at 6 months (Day 180). All treatment groups received doxycycline for the first 30 days. Microfilarial concentrations in 1 ml (mff/ml) blood were determined monthly for 9 months with the modified Knott’s test. A clinical scoring system was employed for each dog enrolled in the study based on thoracic radiography and cardiac ultrasound (CU) examinations performed at Day − 15 (before treatment) and at Day 180. Results Results from the present study suggest that the majority of dogs from all treatment groups became antigen negative, as evaluated at Day 270: 9/10 dogs (90.0%) from G1, 6/10 dogs (60.0%) from G2 and 8/10 dogs (80.0%) from G3. Improvement of radiographic alterations was observed in all treatment groups, and almost all dogs were cleared of pulmonary abnormalities by 6 months from the beginning of treatment (P = 0.000). Cardiac ultrasound examination showed a progressive improvement of cardiac function in a limited number of animals (4/30). Conclusions The combination of doxycycline and three different formulations of moxidectin leads to antigen-negative status in naturally infected dogs.

Abstract Background Parasites that infect cats include protozoa, helminths and arthropods, many of which are transmissible to humans. Effective control relies on a good knowledge of parasite distribution and the risk factors for infection. The present study was aimed at evaluating the prevalence of major feline parasites in Italy and the risk factors associated with their occurrence. Methods Over a 12-month study period, feces, hair and ectoparasites from naturally infected cats from feral colonies, shelters and private households were analyzed at 13 study centers across Italy. Samples from these cats (n = 987) were analyzed at all centers using the same diagnostic methods. Prevalence values and risk factors were evaluated statistically for the identification of predictors of risk. Results The overall prevalence of gastro-intestinal and broncho-pulmonary (BP) nematodes was 35.9% (354/987). Toxocara cati was the most prevalent species (253/987; 25.6%), followed by Ancylostomatidae (98/987; 9.9%). Among BP nematodes, Aelurostrongylus abstrusus was the most common (76/987; 7.7%). Approximately 35.7% (352/987) of the study population was infested by ectoparasites, of which the most common were fleas (29.4%, 290/987), followed by ear mites Otodectes cynotis (9.8%, 97/987). Predictors of risk for parasite infection included age, a predominantly or exclusively outdoor lifestyle, geographic area and lack of antiparasitic treatment. Conclusions Both ecto- and endoparasites are still common in cats throughout Italy, many of them being of zoonotic concern and vectors of pathogens to humans. Given the presence of parasites throughout the entire study period, year-round treatment should be considered. Furthermore, data confirm the need to protect the human–animal bond using proper endo- and ectoparasiticides to reduce the risk of human infection, in application of the One-Health concept. Graphical abstract

Abstract Background Cystic echinococcosis (CE) is a worldwide parasitic zoonosis caused by the larval stage of Echinococcus granulosus sensu lato affecting livestock, particularly sheep and goats. However, often this parasitosis is underestimated. For this reason, this study aimed to evaluate the epidemiological features and spatial distribution of CE in sheep and goats slaughtered in a hyperendemic Mediterranean area. Methods A survey was conducted in the Basilicata region (southern Italy) from 2014 to 2019. A total of 1454 animals (1265 sheep and 189 goats) from 824 farms were examined for hydatid cyst detection by visual inspection, palpation and incision of target organs. All the CE cysts were counted and classified into five morphostructural types (unilocular, multiseptate, calcified, caseous and hyperlaminated). Molecular analysis was performed on 353 cysts. For spatial analysis, a kriging interpolation method was used to create risk maps, while clustering was assessed by Moran’s I test. Results CE prevalence of 72.2% (595/824) and 58.4% (849/1454) was observed at the farm and animal levels, respectively, with higher values in sheep (62.9%) than goats (28.0%). The liver and lungs were the most frequently infected organs in both sheep and goats. Most of recovered cysts were of the calcified and multiseptate morphotypes. All the isolates were identified as E. granulosus sensu stricto (genotypes G1–G3). Spatial distribution showed a moderate clustering of positive animals. Conclusion The findings of this study can be used to better understand the eco-epidemiology of echinococcosis and to improve CE surveillance and prevention programs in regions highly endemic for CE. Graphical abstract

Abstract Background The worldwide increased difficulty to combat gastrointestinal nematode (GIN) infection in sheep, due to progressing anthelmintic resistance (AR), calls for an enhanced and standardized implementation of early detection of AR. This study provides a snapshot of the current AR status against benzimidazoles and macrocyclic lactones in southern Italy, generated with standardized techniques. Methods On 10 sheep farms, the efficacy of albendazole (ALB) and either eprinomectin (EPR) or ivermectin (IVM) was evaluated based on the faecal egg count reduction test (FECRT) performed with the Mini-FLOTAC. For each tested drug, 40 sheep were rectally sampled at D0 and sampled again 14 days after the treatment (D14). The FECRT was calculated from individual samples and pooled samples which consist of 5 individual samples. Efficacy was classified as ‘reduced, ‘suspected’ and ‘normal’. Coprocultures were set for D0 and D14 faecal samples of each group. From farms with FECR < 95%, an in vitro egg hatch test (EHT) and a follow-up FECRT using fenbendazole (FBZ) were conducted. Results Based on the FECR, high efficacy (from 95.7% to 100%) was observed for ALB and IVM in eight farms (Farms 3–10). On Farm 1 and Farm 2, the efficacy for the macrocyclic lactones was classified as ‘normal’, but ‘reduced’ efficacy was observed for ALB on Farm 1 (FECR = 75%) and ‘suspected’ efficacy on Farm 2 (FECR = 93.3%) with the predominant GIN genus Trichostrongylus followed by Haemonchus at D14. The FEC results of pooled samples strongly correlated with those of individual samples, for FEC at D0 (r s = 0.984; P < 0.0001) and at D14 (rs = 0.913; P < 0.0001). The classifications of efficacy in Farm 1 (FECR = 86.0%) and Farm 2 (FECR = 93.0%) in the follow-up FECRT with FBZ coincide with the main FECRT trial. The in vitro EHT confirmed AR in both farms (Farm 1: 89%; Farm 2: 74%). Conclusions In regions like southern Italy, where the negative impacts from AR have played a minor role, efficient monitoring of AR is important in order to evaluate potential risks and being able to promptly respond with countermeasures.

Abstract Background Faecal egg count (FEC) techniques are commonly used to detect gastrointestinal nematodes (GINs) in cattle and to determine anthelmintic efficacy/resistance through the faecal egg count reduction test (FECRT). Mini-FLOTAC is one of the techniques recommended for a standardised FEC/FECRT of helminth eggs in cattle. However, only one paper evaluated the recovery rate of GIN eggs by Mini-FLOTAC (compared to McMaster and modified-Wisconsin method) in cattle, using only a level of contamination of 200 eggs per gram (EPG) of faeces and using GIN eggs collected from goat faeces to spike faecal samples from cattle. To further study the recovery rate of GIN eggs from cattle faeces, this study was conducted in two laboratories, one in Belgium and one in Italy to evaluate the sensitivity, accuracy, precision and reproducibility of the Mini-FLOTAC and McMaster techniques (at two reading levels: grids and chambers) for the detection of GIN eggs in spiked bovine faecal samples. Methods In both countries, spiked cattle faecal samples with five different levels of egg contamination (10, 50, 100, 200 and 500 EPG) of GINs were used. The study was performed in both laboratories by the same expert operator and using the same standard operating procedures (SOPs) for the Mini-FLOTAC and McMaster techniques. Sensitivity, accuracy and precision were calculated for each technique and for each level of contamination. Statistical analyses were performed to evaluate differences in performance between the two techniques. Results Mini-FLOTAC had a higher sensitivity (100% at all EPG levels for Mini-FLOTAC vs 0–66.6% for McMaster chambers and grids at levels

Abstract Background Schistosomiasis and food-borne trematodiases are not only of major public health concern, but can also have profound implications for livestock production and wildlife conservation. The zoonotic, multi-host nature of many digenean trematodes is a significant challenge for disease control programmes in endemic areas. However, our understanding of the epidemiological role that animal reservoirs, particularly wild hosts, may play in the transmission of zoonotic trematodiases suffers a dearth of information, with few, if any, standardised, reliable diagnostic tests available. We combined qualitative and quantitative data derived from post-mortem examinations, coprological analyses using the Mini-FLOTAC technique, and molecular tools to assess parasite community composition and the validity of non-invasive methods to detect trematode infections in 89 wild Hubert’s multimammate mice (Mastomys huberti) from northern Senegal. Results Parasites isolated at post-mortem examination were identified as Plagiorchis sp., Anchitrema sp., Echinostoma caproni, Schistosoma mansoni, and a hybrid between Schistosoma haematobium and Schistosoma bovis. The reports of E. caproni and Anchitrema sp. represent the first molecularly confirmed identifications for these trematodes in definitive hosts of sub-Saharan Africa. Comparison of prevalence estimates derived from parasitological analysis at post-mortem examination and Mini-FLOTAC analysis showed non-significant differences indicating comparable results between the two techniques (P = 1.00 for S. mansoni; P = 0.85 for E. caproni; P = 0.83 for Plagiorchis sp.). A Bayesian model, applied to estimate the sensitivities of the two tests for the diagnosis of Schistosoma infections, indicated similar median posterior probabilities of 83.1% for Mini-FLOTAC technique and 82.9% for post-mortem examination (95% Bayesian credible intervals of 64.0–94.6% and 63.7–94.7%, respectively). Conclusions Our results showed that the Mini-FLOTAC could be applied as an alternative diagnostic technique for the detection of the zoonotic S. mansoni and other trematodes in rodent reservoirs. The implementation of non-invasive diagnostics in wildlife would offer numerous advantages over lethal sampling methodologies, with potential impact on control strategies of zoonotic helminthiases in endemic areas of sub-Saharan Africa and on fostering a framework of animal use reduction in scientific practice.

Abstract Background Faecal egg counts (FEC) and the FEC reduction test (FECRT) for assessing gastrointestinal nematode (GIN) infection and efficacy of anthelmintics are rarely carried out on ruminant farms because of the cost of individual analyses. The use of pooled faecal samples is a promising method to reduce time and costs, but few studies are available for cattle, especially on the evaluation of different pool sizes and FECRT application. Methods A study was conducted to assess FEC strategies based on pooled faecal samples using different pool sizes and to evaluate the pen-side use of a portable FEC-kit for the assessment of FEC on cattle farms. A total of 19 farms representing 29 groups of cattle were investigated in Italy and France. On each farm, individual faecal samples from heifers were collected before (D0) and two weeks after (D14) anthelmintic treatment with ivermectin or benzimidazoles. FEC were determined individually and as pooled samples using the Mini-FLOTAC technique. Four different pool sizes were used: 5 individual samples, 10 individual samples, global and global on-farm. Correlations and agreements between individual and pooled results were estimated with Spearman’s correlation coefficient and Lin’s concordance correlation coefficients, respectively. Results High correlation and agreement coefficients were found between the mean of individual FEC and the mean of FEC of the different pool sizes when considering all FEC obtained at D0 and D14. However, these parameters were lower for FECR calculation due to a poorer estimate of FEC at D14 from the faecal pools. When using FEC from pooled samples only at D0, higher correlation and agreement coefficients were found between FECR data, the better results being obtained with pools of 5 samples. Interestingly, FEC obtained on pooled samples by the portable FEC-kit on-farm showed high correlation and agreement with FEC obtained on individual samples in the laboratory. This field approach has to be validated on a larger scale to assess its feasibility and reliability. Conclusions The present study highlights that the pooling strategy and the use of portable FEC-kits on-farm are rapid and cost-effective procedures for the assessment of GIN egg excretion and can be used cautiously for FECR calculation following the administration of anthelmintics in cattle.

Abstract Background Giardia spp. is a protozoan pathogen and is the most common enteric parasite of domestic animals and humans. Assays for detecting infection in fecal samples using direct or indirect examinations are important tools for diagnosing the disease. The objective of the present study was to compare the cost-effectiveness of immunoassays and FLOTAC technique for diagnosing Giardia spp. infection in dogs. Results Fecal samples from 80 positive stray dogs were tested for the presence of copro-antigens of Giardia spp. using the direct immunofluorescence assay (IFA), a rapid enzyme-linked immunosorbent assay (ELISA) and the FLOTAC double technique. All methods were performed in accordance with the instructions reported in the original description for each technique. The results showed that ELISA can be run in less time than IFA and almost at the same time of the FLOTAC technique. Among the tests used in this study, FLOTAC had the lowest cost per correct diagnosis, compared with immunoassays. Conclusions The results from this cost-effectiveness analysis, in combination with the sensitivity and specificity of the FLOTAC technique, suggest that the FLOTAC technique can be use in the routine diagnosis of Giardia spp. infection in dogs.

Abstract Background The geographical distribution of ticks on companion animals needs to be monitored to develop and plan effective control measures, as suggested by the European Scientific Counsel on Companion Animal Parasites. The aim of this study was to conduct the first Italian national survey of tick distribution on privately owned dogs. Methods The study was performed over 20 months (February 2016 - September 2017) and involved 153 veterinary practices in 64 different provinces covering 17/20 (85%) Italian regions. Participating practitioners were asked to examine five different dogs per month at random and complete a questionnaire for each dog. Differences in tick infestation associated with: sex, age and hair length (long and short); the dog’s habitat (indoor or outdoor/kennel); and the dog’s environment (urban or rural/sylvatic) were evaluated. The attachment site of ticks on the dog was also recorded. Acaricide efficacy was evaluated for the subset of dogs for which complete information on product used, date of sampling and date of last ectoparasiticide treatment was available. Results Of the 3026 dogs examined, 1383 (45.7%) were carrying at least one tick. Overall, 2439 tick samples were collected and a total of 14 tick species identified. Rhipicephalus sanguineus group were the most predominant ticks (63.6%), followed by Ixodes ricinus (30.6%) and I. hexagonus (5.6%). Twenty-four dogs had mixed tick infestations. Long-haired dogs had a higher tick infestation risk as did dogs with outdoor and rural/sylvatic lifestyles. Ticks were located on the head (37.4%), the neck (28.8%), the muzzle (15.5%) and the back (15.3%). A higher prevalence of Rhipicephalus was found in the interdigital spaces (10.8%) compared to Ixodes (0.2%). Finally, ectoparasiticide treatments were found significantly protective against tick infestation, especially orally administered formulations. Conclusions Privately owned dogs in Italy have a high prevalence (45.7%) of infestation with ixodid ticks and this risk varies by dog phenotype and lifestyle.

Abstract Background Analyses of environmental correlates of the composition of gastrointestinal parasite communities in black howler monkeys (Alouatta pigra) have been hindered by inadequate calibration techniques of detection and quantification methods of the parasites. Here we calibrate samples and compare the likelihood of parasite detection using two flotation techniques, FLOTAC and Mini-FLOTAC, and compare flotation solution, preservation method and dilution ratio for egg detection and counts of the most common parasites (Controrchis spp. and Trypanoxyuris spp.) in howler monkeys. Results For samples preserved in 5% formalin, the Mini-FLOTAC technique was the best option for qualitative and quantitative copro-microscopic analysis. This technique displays an 83.3% and 100% detection of Controrchis spp. and Trypanoxyuris spp. infections, respectively. For the trematode Controrchis spp., more eggs per gram of feces (EPG) were recorded with the flotation solution (FS) #7 (zinc sulfate; specific gravity SG = 1.35) at 1:20 and 1:25 dilution than other methods. By contrast, for the nematode Trypanoxyuris spp., the best results were recorded with FS1 (sucrose and formaldehyde; SG = 1.20) at 1:10 dilution. Conclusions We recommend the Mini-FLOTAC technique for general use with parasite analysis on frugivore/folivores like the howler monkey, especially if many samples are analyzed. The technique has a high detection rate and the best EPG counts, allowing the qualitative and quantitative analysis of parasite load among the species or populations without the need for specialized equipment.

Abstract Background The cat flea, Ctenocephalides felis, is the most prevalent flea species detected on dogs and cats in Europe and other world regions. The status of flea infestation today is an evident public health concern because of their cosmopolitan distribution and the flea-borne diseases transmission. This study determines the spatial distribution of the cat flea C. felis infesting dogs in Spain. Using geospatial tools, models were constructed based on entomological data collected from dogs during the period 2013–2015. Bioclimatic zones, covering broad climate and vegetation ranges, were surveyed in relation to their size. Results The models builded were obtained by negative binomial regression of several environmental variables to show impacts on C. felis infestation prevalence: land cover, bioclimatic zone, mean summer and autumn temperature, mean summer rainfall, distance to urban settlement and normalized difference vegetation index. In the face of climate change, we also simulated the future distributions of C. felis for the global climate model (GCM) “GFDL-CM3” and for the representative concentration pathway RCP45, which predicts their spread in the country. Conclusions Predictive models for current climate conditions indicated the widespread distribution of C. felis throughout Spain, mainly across the central northernmost zone of the mainland. Under predicted conditions of climate change, the risk of spread was slightly greater, especially in the north and central peninsula, than for the current situation. The data provided will be useful for local veterinarians to design effective strategies against flea infestation and the pathogens transmitted by these arthropods.

Abstract Background Kennel dogs are at the high risk of infections with intestinal and extra-intestinal helminths. Therefore, regular parasitological surveillance, appropriate treatment strategies and high quality standard of hygiene are required to guarantee the health and welfare of kennel dogs. The aim of the present study was to evaluate the efficacy of helminth control in kennels using different broad-spectrum anthelmintics that are routinely used in canine veterinary practice. Particular attention was given to the field efficacy and ease-of-use of each product. Methods The study was conducted in 3 public kennels in the Campania region (southern Italy). Eighteen boxes from each of the three kennels were selected for treatment based on faecal egg counts (FECs) at Day −30. The treatments were conducted using tablets containing combinations of: i) milbemycin oxime and praziquantel (Milbemax®); ii) pyrantel embonate, febantel and praziquantel (Drontal Plus Flavour®), and; iii) pyrantel pamoate, oxantel pamoate and praziquantel (Nemex® POP). All dogs were treated on Day 0 and sampled on Days 0, 7, 14 and 21 for copromicroscopic analyses. The FLOTAC dual technique on pooled samples was used with efficacy determined by reduction in FECs. Results At Day −30 intestinal nematodes (hookworms, Toxocara, Trichuris) and cestodes (Dypilidium caninum) were found. Milbemax® showed 100% efficacy against all the helminths in all the kennels. Drontal Plus Flavour® was 100% effective against hookworms in all the kennels but gave lower efficacy against T. canis (range = 97.1-100%) and T. vulpis (range = 95.6-100%). Nemex® POP was also 100% effective against hookworms in all kennels but less effective against T. canis (range = 95.7-100%) and T. vulpis (range = 95.7-100%). All three drugs were 100% effective against D.caninum. Conclusions Milbemax®was fully effective against all the helminthes and was palatable and ease-of-use for all the dogs. It is the optimum choice for the treatment and prevention of helminth infections in kennel dogs under field conditions.

Abstract Background The most common intestinal nematodes of dogs are Toxocara canis, hookworm and Trichuris vulpis. The present study was aimed to validate a new copromicroscopic technique, the Mini-FLOTAC and to compare its diagnostic efficiency and sensitivity with four other copromicroscopic techniques: direct smear, tube flotation, Wisconsin method and the FLOTAC dual technique. Findings Two experiments were performed. In the first, faecal positive samples collected from 59 stray asymptomatic dogs, of which 21 were naturally infected with ancylostomidae, 13 naturally infected with T. canis and 25 naturally infected with T. vulpis were used to validate the Mini-FLOTAC technique. The second experiment was performed on faecal samples randomly selected from 38 stray asymptomatic dogs to compare the diagnostic efficiency and sensitivity of the different techniques. Samples were fixed with 5% formalin; sodium chloride and zinc sulphate were used for flotation solutions because they performed best for detecting and quantifying intestinal nematode eggs in dogs. Mini-FLOTAC and FLOTAC were the most efficient and sensitive techniques and they gave higher EPG and higher numbers of positive samples in both the experiments, for all three parasites. Conclusions As Mini-FLOTAC does not require centrifugation it is a very promising technique for counting helminth eggs in dog faeces.

Abstract Background Copro-parasitological diagnosis is still a challenge in management of helminth infections at individual and community levels in resource-limited settings. The aim of our study was to compare the performance of three quantitative techniques: Kato-Katz, McMaster and Mini-FLOTAC methids. The study was carried out in Oran, Northern Argentina. Methods 200 schoolchildren were enrolled to provide a single stool sample, which was tested for helminth infections with Kato-Katz, McMaster and Mini-FLOTAC methods. The Mini-FLOTAC was performed with two flotation solutions (FS2 saturated saline and FS7 zinc sulphate). Preparation and reading time for each of the three methods was calculated both when processing single and multiple samples. Results Out of 193 schoolchildren examined, 40% were positive for any helminth infection by any method; the most prevalent was Hymenolepis nana (23%) followed by Ascaris lumbricoides (17%) and a third group of less prevalent helminths: Enterobius vermicularis, Trichuris trichiura and hookworms (11% all together). Mini-FLOTAC FS2 was more sensitive than FS7 for H. nana (93% vs 78%) and for other helminths (85% vs 80%), whereas FS7 was more sensitive for A. lumbricoides (87% vs 61%). Kato-Katz method was more sensitive than McMaster method for A. lumbricoides (84% vs 48%) and for other helminths (48% vs 43%) except for H. nana (49% vs 61%). As for egg counts, Mini-FLOTAC FS2 reported 904 eggs per gram of faeces (EPG) for H. nana (vs 457 with McMaster and 111 with Kato-Katz) and 1177 EPG for A. lumbricoides (vs 1315 with Kato-Katz and 995 with McMaster); FS2 detected the highest EPG for both H.nana and A.lumbricoides (904 vs 568 and 1177 vs 643 respectively), the differences were not statistically significant. The technique feasibility was calculated: Kato-Katz mean time was 48 minutes/sample, Mini-FLOTAC 13 minutes/sample and McMaster 7 minutes/sample. However, especially for Kato-Katz and Mini-FLOTAC, the mean time (min/sample) decreased significantly when processing multiple samples. Conclusions Mini-FLOTAC is a promising technique for helminth diagnosis, it is more sensitive than Kato-Katz and McMaster for H. nana and as sensitive as Kato-Katz and more sensitive than McMaster for A. lumbricoides identification. Egg counts differences although relevant, did not reach statistical significance.