Abstract

Urinary house soiling (periuria) in the home is a common but serious behaviour problem in cats. Although many specific risk factors and triggers have been postulated, their importance is largely unknown. This study assessed: (1) the significance of purported risk factors for periuria as well as specifically marking and latrine behaviour in the home; (2) the specificity and sensitivity of signs commonly used to differentiate latrine and marking behaviour. Owner responses to an internet survey (n = 245) were classified into three groups: control, marking and latrine behaviour, along with 41 potential risk factors and 15 predictors used to diagnose marking and latrine problems. Univariate statistical analyses and non-parametric tests of association were used to determine simple associations. In addition the sensitivity and specificity of four cardinal signs (posture to urinate, attempt to cover soiled area, surface chosen and volume of urine deposited) were calculated. Significant potential risk factors were: age (marking cats were older than the other two groups); multi-cat household (increased risk of marking and latrine behaviours); free outside access and cat flaps in the house (higher frequency of marking); outside access in general (lower prevalence of latrine behaviour); defecation outside the litter box (higher frequency of latrine behaviour); a heavy dependence by the cat on its owner (lower frequency of latrine behaviour) and a relaxed personality (lower risk of marking behaviour). Litterbox attributes and disease related factors were not significant. Individual cardinal signs were generally not good predictors of diagnosis. This study challenges the poor quality of evidence that has underpinned some of the hypotheses concerning the causes of periuria in cats. The results, in particular, highlight the general importance of the social environment, with the presence of other cats in the household, the cat-owner bond and personality related factors, alongside factors like the use of a cat flap which might also alter the social environment, all implicated as significant risk factors. While the physical environment may be important in specific cases, it seems this is less important as a general risk factor. The findings quantify the risk of misdiagnosis if a single sign is considered sufficient for diagnosis.