Cohort Profile: The ‘Bristol Cats Study’ (BCS)–a birth cohort of kittens owned by UK households

This is the authors' abstract. We don't add key points for this paper.

Longitudinal health-related data among domestic cats represents an important information deficit in veterinary literature. Cats are popular household pets in the UK, with approximately 23% of households in the UK owning one or more cats.1 In 2015, the approximate cost of caring for a cat during its lifetime (median duration = 14 years, interquartile range 9.0–17.0)2 was estimated to be approximately £17 000, excluding veterinary costs associated with treating ill or injured cats.3 Commonly diagnosed health problems affecting pet cats of all ages include periodontal disease (68% of cats of all ages)4 and overweight/obesity (11.5–38% of cats of all ages).5,6 Cognitive dysfunction, chronic kidney disease (CKD) and hyperthyroidism are among the most prevalent health conditions diagnosed in cats over the age of 10 years, with reported prevalences within this age group (≥ 10 years) of approximately 36%,7 10%3 and 10%,8 respectively. Among a large sample (n = 3309) of cats attending primary care veterinary practices in England, trauma (including road traffic accidents) was the most commonly attributed (47%) recorded cause of mortality in young (< 5 years) cats, and renal disorder (including CKD) was the most commonly attributed cause of mortality (14%) among cats aged 5 years or older within this sample.2 Unwanted behaviours such as inappropriate elimination (toileting inside the house, but outside the litter box) and aggression towards people and/or other cats are common reasons for which owners seek help from pet behaviourists9 or cite as reasons for relinquishment of pet cats to rehoming organizations.10 The ‘Bristol Cats Study’ (BCS) cohort was set up with the aim of collecting data prospectively from owners and veterinarians (with owner permission) that could be used to advance veterinary knowledge of these and other common feline disorders through examining possible links between owner- and/or veterinary-reported disorders and environmental exposures. To date, publications based on data collected from the BCS cohort include analysis of factors associated with owner reports of lower urinary tract signs (LUTS) (i.e. straining/apparent difficulty urinating, passing blood when urinating and/or vocalizing (e.g. meowing) before or during urination),11 owner-reported road traffic accidents (RTA)12 involving their cat, and overweight/obesity13 (defined as body condition score 4 or 5 on a 5-point scale). These conditions were hypothesized to often not result in presentation of the cat to a veterinarian (LUTS and fatal RTAs) or not be recorded in the veterinary clinical notes (overweight/obesity): hence the value of the BCS owner-reported data. Examples of hypotheses that have been tested, numbers of cats with data available for analyses and statistical power calculations, are summarized in Supplementary Table A, available as Supplementary data at IJE online. Other analyses that are under way include a nested case-control study of risk factors for gingivitis in cats aged 3–4 years and testing fecal samples for parasites.14 Analyses of other outcomes are envisaged and buccal swabs are being collected to facilitate future prevalence studies, subject to identification of variants, sufficient statistical power and funding. The long-term aims of the BCS are to use multivariable regression models to quantify the strength of associations between environmental exposures and common disorders of cohort cats in senior (11–14 years) and geriatric (>15 years) life stages.15 Details of hypotheses to be tested, and the numbers of cats estimated to be available for future analyses of CKD, hyperthyroidism and cognitive dysfunction, are available in Supplementary Table A. Peer-reviewed veterinary research publications often report work conducted at veterinary referral centres, which can result in a ‘referral bias’.16,17 Data collected from first-opinion practices18,19 have other limitations, including a lack of information on lifestyle and limitations intrinsic to veterinary record keeping. In addition, some animals may attend multiple veterinary practices, or move to a practice that is not participating in a particular study. Sometimes animals with a disorder might not be presented to a veterinarian,20 and an estimated 13.6% of UK pet cats are not even registered with a veterinary practice.21 Consequently, prevalence estimates reported from studies using veterinary practice/referral centre data might under-or overestimate the prevalence of diseases if disease prevalence varies according to veterinary registration status. Prospective longitudinal studies offer a number of advantages over cross-sectional and retrospective studies. Cross-sectional studies provide results based only on a point-in-time sample of the population, whereas longitudinal studies allow the critical evaluation and following over time of events preceding the outcome under investigation. The relatively long lifespan of UK pet cats dictates the need for longitudinal studies to better understand benefits of preventive medicine on health outcomes in later life. The original study team of Jane Murray and Tim Gruffydd-Jones (both at the University of Bristol) were aware of the impact and value of the Avon Longitudinal Study of Parents and Children (ALSPAC)22 that was being run by researchers at the University of Bristol. Following a series of meetings with one of the Directors of the ALSPAC study, and ethical approval from the University of Bristol (reference: UIN/13/026), the Bristol Cats Study (BCS) was launched in June 2010. The BCS has been set up and maintained with a small budget. Until December 2014, consumables for the study were funded through profits of Continuing Professional Development (CPD) courses for veterinarians run at the University of Bristol. Cats Protection (the United Kingdom (UK)'s leading feline welfare charity) has funded staff time (Jane Murray s post) since the study launch. Since December 2014, The WALTHAM Centre for Pet Nutrition (a division of Mars Petcare) has funded the study administrator post. Additional funding that has been used to support research students and staff working with BCS data, and to cover some of the consumable costs, has been obtained from the Biotechnology and Biological Sciences Research Council, British Small Animal Veterinary Association Petsavers, The Langford Trust for Animal Health & Welfare, and Zoetis. This report describes recruitment and the first 6 years of follow-up of the BCS cohort. The intention is that follow-up of the cohort will continue for the lifetime of the enrolled cats. Identifying environmental risk factors that can be monitored or avoided could lead to better management of feline health by owners, and hence improved health and welfare for cats. Cat owners (themselves aged 18 years or more) could register 8–16-week-old kittens with the BCS between June 2010 and December 2013 (inclusive). During 2010, registration of pet cats was limited to owners living in the Bristol (‘BS’) postcode region. This geographical restriction was imposed to facilitate visits from researchers to participating households (e.g. to validate questionnaire data and/or to collect more detailed data) and to aid engagement with participating cat owners through BCS conferences/open days. At the end of 2010, only 146 cats had been recruited, so registration was extended across the UK. Advantages of recruiting over a larger area can include access to a wider range of environments (rural, suburban, urban, metropolitan), demographics and socioeconomic strata, in addition to increasing the sample size. To date, analyses11–13,20,23 have not incorporated geographical restriction of cohort (pre/post 2011) as a potential explanatory variable, but future analyses can assess whether geographical restriction is a potential confounder. During 2010, recruitment was limited to just one kitten per household (requested to be the kitten with the name that was first alphabetically) to avoid having to account for clustering at the household level in analysis; however, multiple kitten registration per household was allowed from January 2011 onwards to increase the size of the cohort. Inclusion of more than one cat per owner (or household) in analyses introduces the potential of hierarchical clustering of data. Two-level random intercept models will be used to account for this clustering through assigning level 1 to cat identification and level 2 to owner identification. Cat owners were recruited into the BCS using a variety of advertising methods. These included: asking veterinarians to add BCS leaflets to kitten packs issued to owners when kittens were presented for vaccination; posters displayed in veterinary practices, pet shops and libraries; and advertising through websites used by cat owners (e.g. animal welfare organization websites), in publications aimed at veterinarians and cat owners and via social media (Facebook and Twitter). There are no plans to continue or extend the recruitment of cats to the study. No national database or compulsory registration of pet cats exists in the UK, hence we cannot provide accurate details regarding the eligible sample. In 2011, two studies estimated that there were between 10 114 764 (95% confidence intervals 9 138 603–11 090 924)1 and 11 916 000 pet cats owned by households in the UK.3 Evidence of association between some household factors (e.g. presence of one or more dogs in the household, presence of children aged 11–15 years, highest level of qualification achieved by household members, and age and gender of householders) and cat ownership have been reported,24–26 hence comparison of BCS cohort demographic data with UK census data is likely to be meaningful. A comparison of cat, owner and household demographic data obtained from the BCS cohort and from three other sources of UK pet cat demographic data is provided in Supplementary Table B, available as Supplementary data at IJE online. Data from the BCS cohort were obtained from self-completed questionnaires (see Supplementary data available at IJE online); some data were missing. As self-selected participants in a feline cohort study, compared with the eligible population, the BCS cohort is hypothesized to have a higher proportion of cats that are pure-bred, registered with a vet, insured and microchipped. However, despite potential differences between the demographics of the cohort and the wider population of cats in the UK, these differences are more likely to lead to biased measures of incidence than biases in effect measures. Self-completed questionnaires were completed by owners at registration, when their cats were aged 2–4 months [Questionnaire 1, (Q1)], and have subsequently been issued to owners within about 2 weeks of their cats reaching the following ages: 6 months [Questionnaire 2, (Q2)], 12 months [Questionnaire 3, (Q3)], 18 months [Questionnaire 4, (Q4)], 30 months [Questionnaire 5, (Q5)], 4 years [Questionnaire 6, (Q6)] and annually thereafter. By May 2016, all cats remaining in the study had reached 30 months of age. Final response rates presently are available for Q1–Q5 for the 2203 BCS cohort cats. The number and percentage of questionnaires (Q1–Q5) completed by owners of cohort cats, and the number of cats leaving the study between ages at which questionnaires were issued, are summarized in Figure 1. Numbers of questionnaires completed from the first five questionnaires that were issued to owners of cats registered on the Bristol Cats Study (BCS) when their cat(s) reached specific ages (questionnaires 1-5 (Q1-5)). The numbers of cats leaving the study at each time point are also provided. Questionnaires were completed for 87% of the cats enrolled in the study (and still participating in the study), at age 6 months, and for 79%, 72% and 60% of the cohort at age 12, 18 and 30 months, respectively. The proportions of cats retained at age 6, 12, 18 and 30 months were 95%, 90%, 86% and 82%, respectively. Considering only owners who still had cats registered in the study, questionnaire response rates remained above 72% for the fifth questionnaire which was completed when cats were aged 30 months. Reasons for leaving the study are discussed later and include deceased and missing cats and owners who withdrew from the studies for a variety of reasons. The BCS uses a variety of engagement strategies to maximize retention of owners in the study. Despite these strategies, some cats inevitably will be lost to the study, for example due to mortality or because they have ‘gone missing’. A summary of the reasons reported by owners or recorded by the study team for cats leaving the study to date (27 May 16) is presented in Table 1. The most common reasons for loss to follow-up include cat death from a road traffic accident (21.7%), loss of contact with the owner (20.6%) and owner electing to leave study (20.4%). Owners also reported other reasons for leaving the study, which are described in Table 1. Reasons for cats having left the Bristol Cats Study (27 May 2016) Reasons for cats having left the Bristol Cats Study (27 May 2016) Multivariable logistic regression models were used to test for associations between specific owner and cat variables and cats lost to follow-up, excluding deceased and missing cats, from the study between recruitment and age 18 months (Q4). Loss to follow-up was more likely for cats with owners aged less than 44 years, without a degree or postgraduate qualification, with a household income of less than £30 000 and/or with children in the household at time of Q1 completion. No evidence of association was found for household tenure, source of cat, sex or breed of cat, or presence of other cats in the household, with the likelihood of having dropped out of the study by 18 months.27 This information is useful when assessing potential effects of retention bias on results obtained from the BCS data, in addition to helping to direct future retention strategies. Areas of data collection, which includes owner-completed questionnaires, data collected from veterinarians (body condition scores, oral health scores and clinical notes) and non-invasive sampling carried out by owners is presented in Supplementary Table C, available as Supplementary data at IJE online. Visits to the homes of convenience samples of BCS cat owners have been conducted to validate some owner-reported data (e.g. weight of dry food fed to cats) and oral health scores assigned by the cat’s veterinarian, and to obtain more detailed data related to diets and feeding practices. Links to electronic versions of questionnaires 1–7 (Q1–Q7) are available in the Supplementary data at IJE online. Since December 2012, an annual Christmas card mailed to owners of BCS cats has included oral health and body condition score cards (available as Supplementary dataIJE online) along with pre-paid return envelopes. Owners are requested to take these cards with them to their cat’s annual vaccination and health check, for completion by their veterinarian and return by the owner. On 27 May 2016, one or more sets of veterinary-completed oral health and body condition score cards had been returned by the owners of 864 cats. Two sets of cards had been received for 470 cats, three sets received for 181 cats and four sets received for 29 cats. However, in some instances there were missing data on one of the cards returned, for example if the veterinarian had been unable to examine the mouth of the cat. Supplementary Table D (available as Supplementary data at IJE online) provides details of the broad areas of data collected by owner-completed questionnaires within the BCS to date, and outlines feline samples submitted by owners and data collected from veterinarians of study cats that have been, or are being, collected for cats registered with the BCS. Key findings of the study to date include identifying a strong association between intended and actual age of neutering, and prevalences of obesity and lower urinary tract signs (LUTS) within the cohort. Despite 4 months being the recommended age in the UK for neutering pet cats28,29 (to reduce the risk of unplanned pregnancies), only 14.1% of the cohort were neutered at (or before) 4 months of age, whereas at 12.5–13 months of age, 73.5% were reported to have been neutered at (or before) 6 months of age.21 If owners of cats enrolled on the BCS are more likely than cat owners who are not enrolled on the BCS to be engaged with the veterinary profession (Supplementary Table B), then these results provide strong evidence that there is scope to increase awareness among pet owners and the veterinary profession regarding the recommended age of neutering. A total of 7% of owners reported their cats to be overweight or obese (defined as a body condition score of 4 or 5 on a five-point scale) at around 1 year of age. The two factors found to be independently associated with an increased risk of feline obesity at age 12.5–13 months of age were cats with restricted or no outdoor access, and cats fed dry food as the only or major (> 50%) type of food in their is a health in just as is in from studies that obese cats are more likely to more likely to conditions and more likely to compared with being overweight also may have effects on health in to obesity in a number of in a follow-up study of a feline obesity prevalence study by of overweight cats had weight 4 years just had lost weight and remained from the BCS cohort may be used by owners to help reduce the likelihood of cats overweight or The most owner-reported in cats, as cats owners had the cat urinating and reported one or more of the following straining/apparent difficulty urinating, passing blood when urinating and/or vocalizing (e.g. meowing) before or during increased risk at 18 months of age was associated with cats with an and for cats that had received a in between the ages of 12 and 18 The prevalence of owner-reported in cohort was higher than reported in studies of cats attending veterinary and the risk of in that into the effects of (and on prevalence of is publications based on data collected from the cohort have been by details of these publications are available on the Bristol Cats Study The of the BCS cohort are the large sample size and retention of owners in the study over 72% of cats enrolled in the study had their questionnaires completed by their owners In researchers working on a longitudinal study of dogs in the reported that only of owners were in the study when dogs were aged or The of detailed household, and management data variables associated with these factors to be included in risk in to studies based on data obtained from veterinary of clinical signs of problems that are not presented to veterinarians for or only presented the clinical signs more chronic or more in hence data from this study can be used to factors with the potential to provide of The of the study are that despite the size of the there is not statistical power to outcomes and/or exposures with a data to clinical signs also be with as there is potential for of owner reports are compared with clinical data collected from veterinary practices, to assess for bias the data are not available via access or via a researchers to and research based on the Bristol Cats Study data are to contact Jane Murray to research and access to data. The Bristol Cats Study (BCS) is the first study of a cohort of kittens in the 2203 pet kittens living in the UK were recruited via their owners to the BCS between June 2010 and December Owners questionnaires when their cats are aged 2–4 months, 6 months, 12 months, 18 months, years, 4 years and annually thereafter. data include cat and owner demographic household data, management data, cat clinical signs of body condition scores, veterinary and preventive Additional data collected from owners to for non-invasive samples samples and body condition scores and oral health scores assigned annually by the veterinarians of study cats and submitted via clinical notes from veterinary practices of study cats. In May 2016, cats and owners were still registered on the of the original cohort of 2203 cats. with researchers from other within and outside the UK are and are Supplementary data are available at IJE online. The BCS has been set up and maintained with a small budget. Until December 2014, consumables for the study were funded through profits of Continuing Professional Development (CPD) courses for veterinarians run at the University of Bristol. Cats Protection (the leading feline welfare charity) has funded staff time (Jane post) since the study launch. Since December 2014, The WALTHAM Centre for Pet Nutrition (a division of Mars Petcare) has funded the study administrator post. Additional funding that has been used to support research students and staff working with BCS data and to cover some of the consumable costs, has been obtained from the Biotechnology and Biological Sciences Research Council, British Small Animal Veterinary Association Petsavers, the Langford Trust for Animal Health & Welfare, and Zoetis. is for and before the of the Bristol Cats owners of cohort cats for their in the study and for their to data. and Jane Murray are funded by and Cats respectively. The following are for their help with study and is for during of the of