VN July 2026

Vetnuus | July 2026 11 Survey content The sections of the questionnaire relevant to the current study included: 1. The first section captured owner demographics, including age, gender, postcode and occupation, and canine demographics, including breed, sex, date of birth and their generation (if the dog was a crossbreed). 2. To collect data on dog behaviour, the Canine Behavioural Assessment and Research Questionnaire (C-BARQ) [14] was used to explore six broad aspects of dog behaviour: training and obedience, aggression, fear and anxiety, separationrelated behaviour, excitability, and attachment and attentionseeking. C-BARQ is a well-established tool that includes 73 questions, which are aggregated during analysis into 12 numerical subscales within the C-BARQ(100) scoring method [5]. To limit the length of the survey and to reduce response dropout, questions under the ‘miscellaneous problem behaviours’ category of the original C-BARQ survey were omitted. These items have previously been shown to be optional and removable from the questionnaire without compromising the reliability or validity of the remaining subscales [1]. 3. To quantify which (if any) resources were used by the owner to help with dog behaviour and training, and whether the owner had ever sought professional advice for their dog’s behaviour, further questions were either adapted from previous surveys published by the current research group [2,27,28] or were newly curated by the same team. These questions were presented as single-choice or multiple-choice with an additional option for free-text responses. 4. To explore the difference between owners’ expectations of behaviour and training before acquisition compared to the reality after acquisition of their dog, two additional multiplechoice questions were included with an option for free-text responses to expand upon their answers. The full study questionnaire also included questions regarding the dog’s health and husbandry that are not reported here. The full questionnaire can be found in the S2 File. Data analysis The questionnaire data were manually cleaned in Microsoft Excel (2013). Respondents were excluded if their responses did not show that they met all of the inclusion criteria. Statistical analysis used IBM SPSS Statistics (V29.0.0.0) software. Responses were combined to represent all three Poodle-type breeds (Toy, Miniature and Standard) as one Poodle group for analysis purposes, to maximise statistical power for comparisons [1]. Descriptive statistics reported frequency and percentage for relevant variables. Median, range and interquartile range (IQR) of dog age (years) were calculated at the date of completion of the questionnaire. Other demographic variables were compared between designer-crossbreeds and their progenitors using chisquare testing. Analysis of variance (ANOVA) was used to compare mean breed scores for each of the 12 C-BARQ subscales. The two-way ANOVA models included a fixed set of additional variables to account for confounding that were selected using an information theory approach [29] developed from other canine welfare studies. The following were included as confounding factors [30,31]: dog age, sex and neuter status, owner gender and age, whether the puppy had been seen with the mother on day of collection (as a proxy of illegal sales [27]), whether the owner worked in the canine sector (given potentially differing experience levels/perceptions of behaviour), whether the respondents was the primary dog owner, first-time ownership and health disorder prevalence (given the emerging relationship between health and behaviour [32] using VetCompass coding [33]. The p-value for statistical significance for all analyses was set at p < 0.05. Results From n = 10,524 responses received, n = 1,113 (10.6%) did not meet all the inclusion criteria, and n = 9 (0.1%) had missing data for some of the inclusion criteria, leaving 9,402 (89.3%) valid responses that met all the inclusion criteria and were included in the final analytic dataset. Dog demographics The n = 9,402 dogs in the final analysis included n = 3,424 (36.4%) designer-crossbreed and n = 5,978 (63.6%) purebred dogs. The designer-crossbreed dog group comprised of n = 1,856 (19.7% of 9,402 dogs) Cockapoos, n = 583 (6.2%) Labradoodles and n = 985 (10.5%) Cavapoos. The purebred dog group comprised of n = 715 (7.6%) CKCS, n = 2,237 (23.8%) Cocker Spaniels, n = 2,099 (22.3%) Labrador Retrievers, n = 352 (3.7%) Miniature Poodles, n = 315 (3.4%) Standard Poodles and n = 260 (2.8%) Toy Poodles, with the three Poodle breeds combined represented by n = 927 dogs (9.9%). All dogs included in the study were required to be 5 years old or under. The overall median age of all dogs in the study was 22 months (IQR: 12–34, range: 9–56). Designer-crossbreeds were significantly younger than purebred breeds (p < 0.001), with the median age of dogs within the designer-crossbreed group being 19 months (IQR: 10–31, range: 1–55), while the median age of dogs in the purebred breed group was 24 months (IQR:13–32, range: 1–56). Of the designer-crossbreeds with data available, n = 1,899 (20.2%) were reported as F1 generation (first-generation cross between two purebred progenitor breeds), n = 225 (2.4%) F2 generation, n = 32 (0.3%) F3 generation and n = 143 (1.5%) multigenerational. The full details of designer-crossbreed generation can be found in Table 1. An F1 (n = 1,899, 20.2%) was the most commonly reported generational type overall across all three designer-crossbreeds. Labradoodles were significantly less likely to be reported as F1 compared to the other two designer-crossbreeds (F1: Cockapoo: n = 1,103, 59.7%; Cavapoo: n = 600, 61.3%; Labradoodle: n = 196, 33.8%; p < 0.001). A total of n = 154 (1.6%) owners reported they did not know their dog’s generation, and the generation of n = 365 (3.9%) designer-crossbreeds was unspecified. >>>12 Leading Article

RkJQdWJsaXNoZXIy OTc5MDU=