Epidemiological study of dogs with diabetes mellitus attending primary care veterinary clinics in Australia

Abstract

Background: The objectives of this study were to establish the prevalence, risk factors and comorbidities/sequelae for diabetes mellitus (DM) in Australian dogs presented to first-opinion veterinary practices.

Methods: Electronic patient records of dogs (n=134,329) attending 152 veterinary clinics during 2017 were sourced through VetCompass Australia. They included 418 dogs with DM; a prevalence of 0.36 per cent (95 per cent CI 0.33 per cent to 0.39 per cent) in Australian dogs attending these veterinary clinics. By comparing with the reference group of rarer breeds and unidentified crossbreeds, multivariable modelling was used to reveal breeds (and their crosses) with significantly higher odds of having DM.

Results: The results revealed that breeds (and their crosses) with significantly higher odds of having DM were Australian terriers (ORs=7.93 (95 per cent CI 2.83 to 22.27)), Siberian huskies (OR=6.24 (95 per cent CI 2.51 to 15.54)), English springer spaniels (OR=5.37 (95 per cent CI 1.48 to 19.53)), West Highland white terriers (OR=4.85 (95 per cent CI 2.55 to 9.25)), miniature schnauzers (OR=3.47 (95 per cent CI 1.16 to 10.35)), all types of poodles (OR=3.41 (95 per cent CI 2.07 to 5.61)), bichon frises (OR=3.41 (95 per cent CI 1.65 to 7.01)), schnauzers (OR=3.18 (95 per cent CI 1.42 to 7.11)) and cavalier King Charles spaniels (CKCS; OR=1.84 (95 per cent CI 1.08 to 3.13)). Breeds with lower risk were German shepherd dogs (OR=0.11 (95 per cent CI 0.01 to 0.84)), golden retrievers (OR=0.09 (95 per cent CI 0.01 to 0.68)) and boxers (no cases identified). Fisher's exact tests showed that labradoodles were diagnosed significantly more often than purebred Labradors (P=0.04) and did not differ significantly from poodles (P=0.81). Cavoodles did not differ significantly from either CKCS (p~1.00) or poodles (P=0.12). Spoodles were significantly less diagnosed than poodles (P=0.003) but did not differ from cocker spaniels (P=0.66). Desexed male dogs had a higher odds of DM than entire male (OR=0.62 (95 per cent CI 0.39 to 0.98)) and desexed female dogs (OR=0.76 (95 per cent CI 0.61 to 0.96)). Comorbidities/sequelae associated with canine DM included suspected pancreatitis (OR 10.58 (95 per cent CI 5.17 to 22.78)), cataracts (OR 9.80 (95 per cent CI 5.65 to 17.35)), hyperadrenocorticism (OR 6.21 (95 per cent CI 3.29 to 11.88)), urinary tract infection (OR 5.09 (95 per cent CI 1.97 to 13.41)) and hypothyroidism (OR 4.10 (95 per cent CI 1.08 to 15.58)).

Conclusions: Breeds at most risk included Australian terriers and Siberian huskies as previously reported, as well as, for the first time, English springer spaniels. In contrast to other populations where there is female predisposition for DM, desexed male dogs in Australia were at increased risk for DM compared with both entire males and desexed females. This predisposition for desexed males to develop DM warrants further investigation.

Keywords: diabetes mellitus; disease surveillance; dogs; epidemiology.