A retrospective cross-sectional survey of dog-owning veterinary clients was
undertaken in 1996 in the three Canadian provinces of New Brunswick, Nova
Scotia, and Prince Edward Island, to generate a population of dogs for futu
re use in a more detailed survey on canine behaviour. The questionnaire was
designed to detect which dogs had or had not bitten a person living in the
same household, and included both demographic and behaviour questions. Twe
nty veterinary clinics were enlisted to administer the questionnaire to the
ir clientele. Data was collected on 3226 dogs, a response rate of 81.4%. Do
gs were predominantly purebred (60.1%) and neutered (71.6%). The Labrador R
etriever was the most commonly reported of 110 breeds. There were slightly
more female than male dogs, and significantly more female dogs were neutere
d (P < 0.001). Questions elicited information about three forms of aggressi
on: growling, possessive aggression, and biting. The reported frequencies o
f aggression problems were significantly associated with age, gender, neute
r status, and breed. Biting behaviour was reported for 15.6% of all dogs. T
he highest frequency of biting was reported for dogs less than 1 year of ag
e. Relative to intact female dogs of at least I year of age, the odds ratio
for having bitten a member of the household was highest for neutered male
dogs (OR: 3.23; 95% CI: 1.83-5.71), followed by neutered female dogs (OR: 2
.13; 95% CI: 1.21-3.75). Similar trends were seen for growling and possessi
ve aggression. Our results indicate that excellent response rates can be ac
hieved in behavioural research by utilising general veterinary practices an
d their clientele, that canine aggression in a household setting is a frequ
ent problem, and that the relationship between neutering and behaviour warr
ants further investigation. (C) 2001 Elsevier Science B.V. All rights reser
ved.