3

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

aetiological diagnosis. Of the 405 cats that died from trauma, 243 (60.0%) were ascribed to road traffic accidents (RTA). Younger and older cats differed markedly in cause of mortality. 10 For cats dying before 5 years of age (n = 516), the most frequent causes were trauma (n = 244; 47.3%), viral infectious disorders (n = 34; 6.6%) and respiratory disorders (n = 23; 4.5%). For cats dying at or after 5 years of age (n = 2,793), the most frequent causes were renal disorder (n = 379; 13.6%), non-specific illness (n = 352; 12.6%), neoplasia (n = 343; 12.3%) and mass lesion disorder (n = 324; 11.6%) (Table 3). For cats dying at or after 5 years (n = 3,360), all risk factors evaluated using univariable linear regression modelling were associated with longevity. Multivariable modelling identified being crossbred, lower bodyweight, neutered and non-insured as associated with increased longevity. After accounting for the effects of the other risk factors, the mean additional longevity for crossbred cats was 0.6 years (95% confidence interval (CI): 0.2 to 1.1, P = 0.008) compared with purebred cats. Compared with entire female cats, neutered female cats had 0.6 years (95% CI: 0.1 to 1.0, P = 0.007) greater longevity and entire male cats had 1.8 years (95% CI: -1.3 to -2.3, P < 0.001) shorter longevity. Increasing bodyweight was associated with decreasing longevity (P < 0.001). The longevity of insured cats was 1.1 years (95% CI: -0.7 to -1.5, P < 0.001) shorter than non-insured cats (Table 4). Adjusting for clustering within veterinary clinics did not materially affect the results. No significant interactions or major departures from normality or homoscedasticity were identified. A relatively low adjusted r 2 value (0.085) suggested that the final model accounted for 8.5% of data variation. 11 Discussion This study of over 4,000 cats attending primary-care veterinary practices aimed to report longevity and mortality values that would generalise well to the overall owned cat population in England and further aimed to explore longevity variation between purebred and crossbred cats. The novel information derived from the study would support evidence-based approaches to advance feline breeding, husbandry and medicine (1-3). The non-normal and bimodal distribution of longevity in the current study (Figure 1) suggested the median as a more appropriate statistic for longevity than the mean because extreme values from non-normally distributed distributions exert disproportionate effects on the mean (27). Few published studies have reported the longevity of domestic cats. A report based on clinical records from 800 Banfield Pet Hospitals in the US suggested an 'average' longevity of 12.1 years for cats but did not specify whether this was a median or a mean value (16). Improved nutrition, health care and management are stated to have increased life expectancy in domestic cats but specific supporting data are limited (2, 15, 17). The current study identified a bimodal longevity distribution peaking in years 1 and 16 that suggested the existence of two distinct sub-populations of cats: those with a propensity for earlier death and cats that survive to an older age (Fig. 1). Longevity is similarly bimodal in dogs (14), indicating that mortality studies should be separated for 12 each longevity group. Of deaths before 5 years of age, almost half (47.3%) resulted from trauma, with the majority resulting from RTA. This compares with just 5.8% of trauma-related deaths in cats dying at or after 5 years of age. The preponderance for RTA-related deaths among younger cats concurs with a UK practice-based study that reported reducing RTA death rates as cats aged (29). A Swedish study of insured cats that died before 12 years of age also identified RTA as important to cat mortality, with an age-standardised mortality rate (ASMR) of 411 per 10,000 cat-years at risk (17). Differing international attitudes to outdoor access for owned cats may affect traumaand RTA-related mortality: over 90% of UK cats have daily outdoor access (30) compared with 80% in Australia (3) and 50-60% in the US (31). The most common causes of mortality at or after 5 years identified in the current study were renal disorder (13.6%), non-specific illness (12.6%), neoplastic disorder (12.3%) and mass lesion disorder (11.6%). The clinical importance of renal disease in older cats is supported by previous reports showing that 15-30% of cats over 15 years old were azotaemic (32). Kidney-and-ureter disorders were the most common cause of mortality in insured cats dying before 12 years of age in Sweden (ASMR: 713 per 10,000 catyears at risk) (17). The frequent identification of physical and biochemical abnormalities in apparently healthy older cats (33, 34) emphasises the value of clinical vigilance and routine health checks to optimise the detection and management of renal 13 disease, especially in older cats (35-37). Although 12% of deaths in the current study were from neoplasia, an equivalent number of deaths were also ascribed to non-specific masses. Although mass lesions could include cysts, inflammation and infection (38), it is possible that many may have been undiagnosed neoplastic disorders, suggesting that neoplasia could account for up to a quarter of deaths in older cats. Neoplastic disorders were reported to have an ASMR of 528 per 10,000 cat-years at risk in insured cats in Sweden but the study was limited by including