Survival analysis to evaluate associations between periodontal disease and the risk of development of chronic azotemic kidney disease in cats evaluated at primary care veterinary hospitals

Rosalie T. Trevejo Banfield Pet Hospital, 18101 SE 6th Way, Vancouver, WA 98683.

Search for other papers by Rosalie T. Trevejo in
Current site
Google Scholar
PubMed
Close
 DVM, MPVM, PhD
,
Sandra L. Lefebvre Banfield Pet Hospital, 18101 SE 6th Way, Vancouver, WA 98683.

Search for other papers by Sandra L. Lefebvre in
Current site
Google Scholar
PubMed
Close
 DVM, PhD
,
Mingyin Yang Banfield Pet Hospital, 18101 SE 6th Way, Vancouver, WA 98683.

Search for other papers by Mingyin Yang in
Current site
Google Scholar
PubMed
Close
 BVMS, MS
,
Catherine Rhoads Banfield Pet Hospital, 18101 SE 6th Way, Vancouver, WA 98683.

Search for other papers by Catherine Rhoads in
Current site
Google Scholar
PubMed
Close
 BA
,
Gary Goldstein College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Gary Goldstein in
Current site
Google Scholar
PubMed
Close
 DVM
, and
Elizabeth M. Lund Banfield Pet Hospital, 18101 SE 6th Way, Vancouver, WA 98683.

Search for other papers by Elizabeth M. Lund in
Current site
Google Scholar
PubMed
Close
 DVM, MPH, PhD

Abstract

OBJECTIVE To examine potential associations between periodontal disease (PD) and the risk of development of chronic azotemic kidney disease (CKD) among cats and determine whether the risk of CKD increases with severity of PD.

DESIGN Retrospective cohort study.

ANIMALS 169, 242 cats.

PROCEDURES Cats were evaluated ≥ 3 times at any of 829 hospitals from January 1, 2002, through June 30, 2013. Cats with an initial diagnosis of PD of any stage (n = 56,414) were frequency matched with cats that had no history or evidence of PD (112,828) by age and year of study entry. Data on signalment, PD, and other conditions potentially related to CKD were extracted from electronic medical records. Cox proportional hazards modeling was used to estimate the association of PD with CKD after controlling for covariates.

RESULTS PD was associated with increased risk of CKD; risk was highest for cats with stage 3 or 4 PD. Risk of CKD increased with age. Purebred cats had greater risk of CKD than mixed-breed cats. General anesthesia within the year before study exit and diagnosis of cystitis at any point prior to study exit (including prior to study entry) were each associated with increased CKD risk. Diagnosis of diabetes mellitus or hepatic lipidosis at any point prior to study exit was associated with decreased CKD risk.

CONCLUSIONS AND CLINICAL RELEVANCE The findings supported the benefit of maintaining good oral health and can be useful to veterinarians for educating owners on the importance of preventing PD in cats.

Abstract

OBJECTIVE To examine potential associations between periodontal disease (PD) and the risk of development of chronic azotemic kidney disease (CKD) among cats and determine whether the risk of CKD increases with severity of PD.

DESIGN Retrospective cohort study.

ANIMALS 169, 242 cats.

PROCEDURES Cats were evaluated ≥ 3 times at any of 829 hospitals from January 1, 2002, through June 30, 2013. Cats with an initial diagnosis of PD of any stage (n = 56,414) were frequency matched with cats that had no history or evidence of PD (112,828) by age and year of study entry. Data on signalment, PD, and other conditions potentially related to CKD were extracted from electronic medical records. Cox proportional hazards modeling was used to estimate the association of PD with CKD after controlling for covariates.

RESULTS PD was associated with increased risk of CKD; risk was highest for cats with stage 3 or 4 PD. Risk of CKD increased with age. Purebred cats had greater risk of CKD than mixed-breed cats. General anesthesia within the year before study exit and diagnosis of cystitis at any point prior to study exit (including prior to study entry) were each associated with increased CKD risk. Diagnosis of diabetes mellitus or hepatic lipidosis at any point prior to study exit was associated with decreased CKD risk.

CONCLUSIONS AND CLINICAL RELEVANCE The findings supported the benefit of maintaining good oral health and can be useful to veterinarians for educating owners on the importance of preventing PD in cats.

Chronic azotemic kidney disease is a common, irreversible, and ultimately debilitating disease of older cats.1 The overall prevalence in the general feline population is estimated to be 1.6%, and prevalence increases with age.2–4 Identification of specific risk factors for CKD can help guide preventive measures for this disease. Investigations of both dogs and cats have yielded somewhat mixed results with respect to associations between CKD and PD. A retrospective cohort study by Glickman et al5 involving canine patients seen at primary care veterinary hospitals from 2002 through 2008 found that the incidence of CKD increased with increasing severity of PD after adjusting for systematic differences in covariates such as age, sex, neuter status, breed, body weight, number of hospital visits, and dental procedures performed. However, a nested case-control study6 of canine patients in the United Kingdom did not find PD to be significantly associated with CKD. A case-control study to identify risk factors associated with CKD in cats found that cats with CKD were more likely to have a prior diagnosis of PD than were cats without CKD.2 A study7 that followed healthy nonazotemic geriatric (> 9 years old) cats found that those with moderate to severe dental disease, classified on the basis of a combined score for the presence and severity of calculus and gingivitis, were at an increased risk of developing CKD, although the timing for the diagnosis of dental disease was unclear.

Numerous studies,8–13 mostly prevalence investigations, of people have found that patients with CKD as evidenced by a glomerular filtration rate below the lower limit of the reference range or a urine albumin-to-creatinine ratio above the upper limit of the reference range are more likely to have PD than are those without evidence of CKD. A retrospective cohort study14 of elderly Japanese patients found that PD was associated with an increase in the cumulative incidence of decreased kidney function (reduced glomerular filtration rate). The exact mechanism by which PD may increase the risk of CKD is not known. However, there is evidence of an increase in markers of inflammation such as serum C-reactive protein and plasma pentraxin-3 in human patients with PD.15–17 A study18 of 38 canine patients with PD found no correlation between serum concentration of C-reactive protein and stage of PD, but did detect a significant decrease in concentrations of the protein following treatment of PD; this result was attributed to alleviation of the systemic inflammatory burden. It has been proposed that subclinical systemic inflammation promotes arthrosclerosis, leading to secondary renal hypoxemia, progressive renal damage, and CKD through localized arterial stenosis, which also reduces cardiac output.5,19,20 Conversely, the adverse effects of uremia have also been hypothesized to increase the risk of PD in feline and human patients with CKD.2,21 This complex interplay between the roles of PD, systemic inflammation, and CKD highlights the importance of well-designed longitudinal studies in estimating the risk of developing CKD among patients with PD.

The primary objective of the study reported here was to evaluate whether cats with PD would be at an increased risk of developing CKD, after controlling for systematic differences in covariates such as age, sex, breed, neuter status, body weight, history of anesthesia, and selected diseases other than CKD in the analyses. A secondary objective was to determine whether the risk of CKD would increase with increasing stage of PD.

Materials and Methods

Animals

The source population from which study cats were selected consisted of all cats evaluated at a Banfield Pet Hospital ≥ 3 times during the study period (January 1, 2002, through June 30, 2013), with the first visit between January 1, 2002, and December 31, 2012, and subsequent visits occurring by June 30, 2013. By the end of June 2013, Banfield had 829 hospitals in 43 states (states without Banfield hospitals were Hawaii, Alaska, Wyoming, North Dakota, West Virginia, Vermont, and Maine), with approximately 20,000 total feline patient visits each week. All visit information at each hospital was entered into a proprietary software program,a and the resulting electronic medical records were uploaded nightly to a central data warehouse for storage.

To be eligible for inclusion in the study, cats were required to have full signalment information, including age, breed, sex, and neuter status (neutered vs sexually intact) on record. Cats with a recorded age > 25 years by the end of the study period were excluded because they occurred infrequently in the database. To minimize the possibility that cats already had the outcome of interest (CKD) when selected for the study, cats were excluded if they had a diagnosis of CKD or acute renal failure, a serum creatinine concentration > 2.2 mg/dL, or a USG < 1.035 prior to or at the time of study entry. Only cats with an initial diagnosis of CKD at or after the time of the third visit during the study period, with a serum or plasma creatinine concentration > 1.6 mg/dL and USG < 1.035 (both measured ≤ 30 days before or after the diagnosis of CKD), were deemed positive for the outcome of CKD. Requiring a minimum number of visits prior to CKD diagnosis helped to ensure that cats did not have preexisting CKD; also, because a cat's first and second visits often occurred in the same week, this ensured a minimal period of medical scrutiny. For the same reason, an interval of ≥ 30 days was required between the first and last visit during the study period for each cat enrolled.

Study design

A retrospective cohort study design was used. The exposed cohort consisted of all cats with an initial diagnosis of staged PD at the time of the first visit during the study period (study entry) that met all study eligibility criteria. No restrictions were implemented with respect to the number of exposed cats enrolled. Cats with a diagnosis of PD that was not assigned a stage were excluded. The unexposed cohort consisted of cats with no recorded diagnosis of PD or any of the following components of PD at any time before the end of the study period: gingivitis, gingival recession, or periodontal pockets.

Cats with PD were frequency matched to cats without PD at a 1:2 ratio on the basis of year of study entry (ie, year of first qualifying visit during study period) and age. This was done to minimize the disparity in age between the groups, considering that cats with PD are typically expected to be older than the general population, and older cats are considered to be at increased risk for CKD.22 A random sample of cats without PD was selected from each calendar year and age group in an effort to mimic the age distribution of cats with PD. The following age categories were used for frequency matching: half-year age groups for cats ≤ 8 years of age, 1-year age groups for cats > 8 through 18 years of age, and all cats > 18 years of age (grouped together because of small sample sizes).

Cats enrolled in the study had been examined by veterinarians and were found to have or not have PD on the basis of gross clinical findings, with or without dental radiographic evidence. Diagnosing veterinarians had access to generally available practice resources that included descriptions and photographs of PD at the various stages to guide them in their assessment. Stage 1 PD was characterized as gingival inflammation with or without gross evidence of dental plaque or calculus, with no gingival attachment loss. Stage 2 was characterized as gingival inflammation with early-stage gingival attachment loss and up to 25% alveolar bone loss (if radiographs were obtained). Stages 3 and 4 were characterized as the presence of any or all of the following: periodontal pockets, bleeding gums, exposed tooth roots, tooth loss, severe halitosis, and > 25% alveolar bone loss (if radiographs were obtained).

The stage of PD at study entry was noted for cats with a diagnosis of PD. Because distinctions between stage 3 and stage 4 PD were only made in the medical record system during the latter part of the study period, these stages were combined for analysis. First, all cats with stage 3 or 4 disease were selected, with the study entry date being the earliest date on which either stage 3 or 4 diagnosis was made (ie, a cat first diagnosed as having stage 3 disease in July 2009, then stage 4 disease in June 2010 would have a July 2009 study entry date; a cat with only stage 4 disease on record would have the first date on which stage 4 disease was recorded as the study entry date). Then, all cats with stage 2 disease were selected, provided they had no previous diagnosis of stage 3 or 4 disease on record. Finally, all cats with stage 1 disease were selected, provided they had no previous diagnosis of stage 4, 3, or 2 disease on record.

The amount of time that each cat contributed to the study (ie, follow-up period) was calculated on the basis of the date of study entry (first qualifying visit during the study period) and the date of study exit (first date of CKD diagnosis for cats that developed CKD or the date of the last visit during the study period for cats that did not develop CKD).

Data collection and preparation

Data on all study variables were extracted from the electronic medical records for each cat. The information collected included age at study entry, body weight recorded closest to time of study entry, breed, sex, neuter status at time of study entry (neutered vs not neutered), the number of visits during the follow-up period, the total number and timing of dental cleanings performed during follow-up period, the total number of FVRCP vaccines received during the patient's lifetime (up to study exit), and whether the cat underwent general anesthesia (for any procedure) during the last year (365 days) of the follow-up period. Data were also obtained on whether any of the following diseases previously hypothesized or found to be associated with either PD or CKD were diagnosed during the patient's lifetime (at any time up to study exit): tooth resorption, stomatitis, FIV infection, hepatic lipidosis, cardiomyopathy, toxoplasmosis, hyperthyroidism, hypertension, cystitis, and diabetes mellitus.6,23–27 No additional data were extracted to validate these diagnoses, which had been made by the attending veterinarians; however, we believed that any degree of misclassification in these diagnoses would be expected to be similar between the exposed and unexposed cohorts. The diagnosis date for each condition of interest was also obtained.

Whether a cat was enrolled in a wellness plan at any point during their follow-up period and, if so, the total duration of plan enrollment during that period were recorded. Wellness plans typically include wellness services and treatments such as vaccinations, deworming, and heartworm testing, as well as unlimited visits to Banfield hospitals. It was considered important to control for this potentially systematic difference in the amount of medical scrutiny that cats underwent, because wellness plan enrollment might have influenced the frequency of visits and, therefore, the likelihood of CKD detection, or could potentially have biased the study findings in other ways.

A review of a subset of the medical records from 161 cats with extreme body weights (< 0.23 or > 15.9 kg [< 0.5 or > 35 lb]) recorded in the weight field indicated that these values were inconsistent with the age or weight information recorded in other areas of the medical record, so they were set as missing values for the analyses. A dichotomous variable for breed was created to control for the potential influence of breed on outcome. Cats described as mixed breed in the medical records or cats for which breed was recorded as domestic shorthair, domestic medium hair, domestic longhair, or other nonspecific breeds were classified as mixed breed. All other cats were classified as purebred. A dichotomous variable was also created for each of the 5 most common purebred cat breeds in the study population, along with 4 additional breeds previously hypothesized to be at higher risk for CKD (Abyssinian, Ragdoll, Burmese, and Birman4,28,29), with the comparison group consisting of all other cats not of that particular breed. The history of a professional dental cleaning during follow-up period was categorized in 2 separate ways: any history of dental cleaning (yes or no) and time since last cleaning (no cleanings, last cleaning > 1 year before study exit, and cleaning done within the year before study exit).

Statistical analysis

The distribution of continuous variables was assessed for normality by construction of histograms, by assessment of skewness and kurtosis, and by the Kolmogorov-Smirnov, Cramer-von Mises, and Anderson-Darling tests. When a nonnormal distribution was evident for continuous variables, results were reported as median and interquartile range. The Kruskal-Wallis test (generalized form of the Mann-Whitney test method) for nonparametric data was used to compare the medians of continuous variables between groups and also to identify potential associations between these variables and risk of CKD. Correlations between pairs of continuous variables were assessed by calculation of the Pearson and Spearman correlation coefficients (r). Differences in the distribution of categorical variables between exposure groups were compared by χ2 analysis. A χ2 analysis was also used to identify unconditional associations between categorical variables and CKD, including the aforementioned conditions hypothesized or previously found to be associated with CKD. For analyses with multiple comparisons, P values were adjusted by means of Bonferroni correction.

A survival analysis approach was used to compare the mean time to development of CKD between cats with and cats without PD. In multivariate modeling to control for covariates, the outcome of interest was time to diagnosis of CKD. Cox proportional hazards models were constructed by forward selection of variables that were identified in unconditional associations to have some evidence of association with the disease (ie, P < 0.20). When there was evidence of strong collinearity between 2 continuous predictor variables (r > 0.6), the variable thought to have the most direct and strongest biological relationship with the outcome was selected for inclusion in the model. Variables with values of P > 0.1 were excluded from the final model, with the exception of sex and neuter status, which were retained in the final model to control for potential confounding, regardless of their association with CKD. Biologically plausible first-order interaction terms for main effects variables included in the final model were evaluated. The risk associated with individual (purebred) cat breeds was evaluated by a separate multivariate model for each breed examined.

The functional form of continuous variables was evaluated according to previously described methods.30 Different forms of these variables (log, square root, and polynomial up to the fifth degree) were evaluated to identify the best functional form for each. Standardized measures of the number of visits and duration that a cat was enrolled in a wellness plan, estimated by dividing these values by the duration of follow-up time for each patient, were similarly evaluated. The proportional hazards assumption (ie, that the effect of the covariate on the HRs was constant over the study period) was assessed separately for each covariate. Graphs of the Kaplan-Meier estimates of the survival function were used to assess the proportional hazards assumption for categorical variables (ie, visual assessment was used to determine whether survival functions for the different PD exposure levels appeared parallel over the follow-up period). In addition, departure from the proportional hazards assumption was assessed for all variables with a transformation of the martingale residuals known as the empirical score process.30 Because most of the variables considered for inclusion in the model failed to meet the proportional hazards assumption, a stratified Cox proportional hazards model was used, with stratification by duration of follow-up time (≤ 350 days, > 350 to 800 days, and > 800 days). The cutoffs of 350 and 800 days were selected to ensure an approximately equal distribution of cats among the 3 strata.

All HRs were reported with 95% CIs and P values for 2-tailed tests. Values of P < 0.05 were considered a strong indication of a systematic influence (not chance variation). Hazard ratios were used to estimate the relative risk of CKD in a group of cats relative to a comparison group. All described statistical analyses were performed with a commercially available computer software program.b

Results

Animals

Between January 1, 2002, and December 31, 2012, 2,383,820 cats were brought to Banfield hospitals, of which 836,275 (35.1%) cats had ≥ 3 visits during that period. Of these, 56,414 cats with PD (ie, exposed cohort) met the remaining inclusion criteria, for an estimated prevalence of staged PD of 7% (56,414/836,275) among potentially eligible cats. The exposed cohort was frequency matched with 112,828 eligible cats without PD (ie, unexposed cohort), for a total study population of 169,242 cats. The overall median follow-up duration for all included cats was 554 days (interquartile [25th to 75th percentiles] range, 258 to 1,080 days). Among 56,414 cats with PD, the stages of PD at study entry were distributed as follows: 24,355 (43.2%) cats had stage 1 disease, 18,848 (33.4%) cats had stage 2 disease, and 13,211 (23.4%) cats had stage 3 or 4 disease. Characteristics of unexposed and exposed (by stage of PD) cats were tabulated; a comparison of the median values (for continuous variables) and the distribution of categorical variables across all 4 exposure categories indicated that there were significant (P < 0.001) differences among the categories for all variables examined (Table 1). Cats with stage 3 or 4 PD were older, weighed less, and had shorter follow-up duration, compared with cats in the remaining PD categories (stage 1 or stage 2). Cats with PD, regardless of stage, were more likely to be enrolled in wellness plans, to have undergone general anesthesia within the year before study exit, or to have had a dental cleaning during follow-up period than were cats without PD. The survival probabilities over the follow-up period (ie, proportion of population remaining past a given time for each exposure group, without controlling for other factors) were summarized (Figure 1).

Figure 1—
Figure 1—

Kaplan-Meier plot of survival probability (ie, proportion of population remaining past a given time) in cats by PD status (categorized as none [n = 112,828; blue line], stage I [24,255; red line], stage 2 [18,848; black line], or stages 3 and 4 combined [13,211; green line]). Vertical marks indicate censored cats.

Citation: Journal of the American Veterinary Medical Association 252, 6; 10.2460/javma.252.6.710

Table 1—

Characteristics of 169,242 cats with (exposed cohort) and without (unexposed cohort) PD evaluated at any of 829 primary care veterinary hospitals.

 No PDPD
Characteristic(n = 112,828)Stage 1 (n = 24,355)Stage 2 (n = 18,848)Stage 3 or 4 (n = 13,211)
Age (y)6.7 (3.6–10.2)4.6 (2.4–8.0)7.0 (4.2–10.3)9.5 (6.4–12.5)
Weight (kg)5.2 (4.2–8.6)5.3 (4.4–6.4)5.4 (4.4–6.5)5 (4–6.3)
Duration of follow-up (d)546 (244–1,083)593 (329–1,089)591 (301–1,115)473 (221–947)
No. of veterinary visits during follow-up period5 (3–7)5 (4–8)5 (4–9)5 (4–8)
No. of FVRCP vaccinations during lifetime2 (0–3)3 (2–5)3 (1–4)2 (1–3)
Female57,788 (51.2)12,094 (49.7)9,073 (48.1)6,369 (48.2)
Neutered105,779 (93.8)23,230 (95.4)18,202 (96.6)12,727 (96.3)
Mixed breed*100,541 (89.1)21,689 (89.1)16,542 (87.8)11,446 (86.6)
Enrolled in a wellness plan at any time during follow-up period55,688 (49.4)18,520 (76.0)13,604 (72.2)8,732 (66.1)
History of dental cleaning during follow-up period30,913 (27.4)12,011 (49.3)10,548 (56.0)8,477 (64.2)
Underwent general anesthesia within year before study exit29,647 (26.3)10,931 (44.9)8,865 (47.0)6,556 (49.6)

Values reported are median (interquartile [25th to 75th percentiles] range) or number (%). Medians were compared between groups with the Kruskal-Wallis test, and differences in the distribution of categorical variables between groups were compared by χ2 analysis; there were statistically significant differences between groups for all variables (P < 0.001).

Mixed-breed cats included 104,570 domestic shorthair, 20,212 domestic medium-hair, and 19,770 domestic longhair cats, as well as 5,666 cats of other mixed or nonspecific breeds.

Descriptive analyses

Of the 169,242 cats in the study population, 3,022 (1.8%) developed CKD during their follow-up period. The median follow-up duration was significantly (P < 0.001) longer for cats with (873 days) than for cats without (550 days) CKD. Cats with CKD differed significantly (P < 0.001 for all comparisons) from those without CKD with respect to the following: they were more likely to have advanced (≥ stage 2) PD; to be female, neutered, or purebred; to have undergone anesthesia within year before study exit; to have a history of dental cleaning during the follow-up period; and to have been enrolled in a wellness plan during the follow-up period. Cats with CKD were also significantly older and weighed less, and they had more veterinary visits, more dental cleanings, and more FVRCP vaccinations than did cats without CKD. Differences were statistically significant (P < 0.001) for all variables examined. A summary of these results was provided (Table 2).

Table 2—

Characteristics of the same cats in Table 1 reclassified on the basis of whether they did or did not develop CKD during the study period.

CharacteristicCKD (n = 3,022)No CKD (n = 166,220)
Age (y)11.2 (8.4–14)6.5 (3.5–10)
Weight (kg)5.1 (4.1–6.4)5.3 (4.2–6.5)
Duration of follow-up (d)873 (422–1,550)550 (255–1,071)
No. of visits during follow-up period7 (4–11)5 (3–7)
No. of FVRCP vaccinations during lifetime2 (1–4)2 (1–3)
PD status
  No PD1,806 (59.8)111,022 (66.8)
  Stage 1332 (11.0)24,023 (14.5)
  Stage 2434 (14.4)18,414 (11.1)
  Stage 3 or 4450 (14.9)12,761 (7.7)
Female1,642 (54.3)83,682 (50.3)
Neutered2,956 (97.8)156,982 (94.4)
Mixed breed2,554 (84.5)147,664 (88.8)
Enrolled in a wellness plan at any time during follow-up period2,030 (67.2)94,514 (56.9)
History of dental cleaning during follow-up period1,617 (53.5)60,332 (36.3)
Underwent general anesthesia within year before study exit1,087 (36.0)54,912 (33.0)

Values for PD status are reported as number (%) of cats with characteristic. Differences were statistically significant for all variables (P < 0.001).

See Table 1 for remainder of key.

Distributions of all health conditions hypothesized a priori to be associated with CKD or PD among cats categorized by PD or CKD status were summarized (Table 3). All of the conditions of interest were significantly more common in cats with PD than in cats without PD, with the exception of hepatic lipidosis. Cardiomyopathy, cystitis, diabetes mellitus, hypertension, hyperthyroidism, and tooth resorption were significantly more likely to be diagnosed in cats with than cats without CKD, whereas hepatic lipidosis was more likely to be identified in cats without CKD. Proportions of cats with a diagnosis of FIV infection, stomatitis, or toxoplasmosis did not differ significantly between cats with and without CKD.

Table 3—

Number (%) of cats in Table 1 with a diagnosis of various conditions previously hypothesized or found to be associated with PD or CKD.

  PD
ConditionNo. of catsYes (n = 56,414)
Cardiomyopathy1,035418 (0.7)
Cystitis20,6027,734 (13.7)
Diabetes mellitus5,2891,932 (3.4)
FIV infection1,516663 (1.2)
Hepatic lipidosis836275 (0.5)
Hypertension357180 (0.3)
Hyperthyroidism8,1002,827 (5.0)
Stomatitis2,5871,628 (2.9)
Tooth resorption4,5852,980 (5.3)
Toxoplasmosis8037 (0.07)

Values of P < 0.05 were considered significant.

Although only 19,024 of 169,242 (11.2%) study cats were purebred, these accounted for 468 of 3,022 (15.5%) CKD cases (P < 0.001). Eighty breeds of purebred cats were represented in this study. The 5 most common purebred cat breeds accounted for most (337/468 [72%]) CKD cases among purebred cats; these included Siamese (119 [25.4%]), Persian (76 [16.2%]), Himalayan (76 [16.2]), Maine Coon Cat (47 [10%)], and Russian Blue (19 [4.1%]). Comparisons of the frequencies of CKD among these breeds as well as Abyssinian, Ragdoll, Burmese, and Birman cats, versus all other cats in the study, were reported (Table 4).

Table 4—

Results of univariate analysis to evaluate risk of CKD among selected breeds of purebred cats relative to the risk for all study cats that were not of each selected breed (referent group).

BreedProportion (%) with CKDRelative risk (95% CI)P value
Siamese119/4,370 (2.7)1.55 (1.29–1.85)< 0.001
Persian76/3,197 (2.4)1.34 (1.07–1.68)0.01
Maine Coon Cat47/2,449 (1.9)1.08 (0.81–1.43)0.62
Himalayan76/2,141 (3.6)2.01 (1.61–2.52)< 0.001
Ragdoll15/864 (1.7)1.03 (0.59–1.61)0.91
Russian Blue19/825 (2.3)1.29 (0.83–2.02)0.26
Abyssinian13/284 (4.6)2.57 (1.51–4.38)< 0.001
Burmese3/244 (1.2)0.69 (0.22–2.12)0.81*
Birman6/129 (4.7)2.61 (1.19–5.70)0.03*

A 2-sided Fisher exact test was used because of small cell values. Values of P < 0.05 were considered significant

Evaluation of correlations among variables indicated that time enrolled in a wellness plan during the study period was positively correlated with number of FVRCP vaccinations (r = 0.63) as well as number of visits during the study period (r = 0.68). The number of dental cleanings was moderately correlated with the number of visits during the study period (r = 0.54) and time enrolled in a wellness plan during the study period (r = 0.51). To address these correlations, visit frequency was selected for evaluation in the multivariate model in favor of the other correlated variables, as it was felt to be a more direct measure of the amount of medical scrutiny and preventive care cats received. Similarly, a history of anesthesia within the year before study exit was correlated with a history of dental cleaning during the same period (r = 0.65). Of 55,999 cats that had undergone anesthesia within the year before study exit, 39,014 (69.7%) had also had a dental cleaning during the same period. The median age of cats that had a dental cleaning was slightly, but significantly, higher than that for cats with no history of dental cleaning (7.3 vs 7.1 years, respectively). To avoid redundancy, history of anesthesia was selected for evaluation in the final model, in favor of history of dental cleaning, because it was more prevalent and had been identified in a previous study2 as a risk factor for CKD.

Multivariate analysis

The final hazard model with a diagnosis of CKD during the study period as the outcome and stage of PD as the main predictor also included sex, neuter status, breed category, body weight (as a fourth-order polynomial), age (as a third-order polynomial), number of visits (as a third-order polynomial), history of anesthesia within the year before study exit, and a diagnosis of cystitis, diabetes mellitus, or hepatic lipidosis at any point during the patient's lifetime up to study exit (Table 5). After adjusting for covariates, the risk of CKD was significantly higher for cats with any stage of PD than that for cats without the condition. The risk for a diagnosis of CKD in cats with stage 3 or 4 PD was 1.5 times that of cats without PD (95% CI, 1.35 to 1.67). The risk of CKD in cats with stage 1 or 2 PD was slightly lower than that associated with stage 3 or 4 disease, but still associated with an increased risk of CKD; the differences in risk among the 3 categories of PD were not statistically significant. The risk of CKD increased significantly with increasing age, with a nearly 40% increase in risk with each year of age (HR, 1.39; 95% CI, 1.23 to 1.56). Cats with a history of general anesthesia within the year before study exit had a significantly higher risk of CKD diagnosis, compared with that for cats with no such history. Cats identified as mixed breed had a lower risk of a CKD diagnosis than did purebred cats. There was an increased overall risk of CKD associated with a diagnosis of cystitis; an interaction term between sex and cystitis was significant, reflecting that the risk of CKD associated with cystitis was significantly higher in females than in males. Neuter status was not associated with the risk of CKD. A lower risk of CKD was observed in cats with diabetes mellitus or hepatic lipidosis, compared with the risk for cats without these conditions.

Table 5—

Results of multivariable Cox regression analysis to identify risk factors for a diagnosis of CKD in the cats represented in Table 1.

VariableHR (95% CI)
PD
  NoneReferent
  Stage 11.33 (1.18–1.49)*
Stage 21.34 (1.20–1.49)*
Stage 3 or 41.50 (1.35–1.67)*
Weight (kg)1.29 (0.88–1.88)
Age (y)1.39 (1.23–1.56)*
Breed category 
PurebredReferent
Mixed breed0.76 (0.69–0.84)*
Neutered 
NoReferent
Yes1.12 (0.88–1.43)
Sex 
MaleReferent
Female0.92 (0.84–1.00)
No. of visits during follow-up period1.01 (0.99–1.03)
Underwent general anesthesia within year before study exit 
NoReferent
Yes1.62 (1.49–1.76)*
Diagnosis of cystitis during lifetime 
NoReferent
Yes1.49 (1.30–1.71)*
Diagnosis of diabetes mellitus during lifetime 
NoReferent
Yes0.80 (0.66–0.97)*
Diagnosis of hepatic lipidosis during lifetime 
NoReferent
Yes0.14 (0.04–0.57)*
Diagnosis of cystitis in females versus males (interaction between sex and cystitis)1.21 (1.04–1.41)*

Risk differs significantly (P < 0.05) from that of referent group.

Fourth-order polynomial.

Third-order polynomial.

Given the association of Siamese, Persian, Himalayan, Abyssinian, and Birman breeds with a diagnosis of CKD in unconditional analysis, these breeds were each evaluated separately in the multivariate model, with the referent group for each of these comparisons consisting of all other study cats. Only the following 3 breeds were found to have a significantly increased risk of CKD after adjusting for other study covariates and applying Bonferroni adjustment for multiple comparisons: Siamese (HR, 1.36; 95% CI, 1.13 to 1.63), Himalayan (HR, 1.44; 95% CI, 1.14 to 1.80), and Abyssinian (HR, 2.34; 95% CI, 1.36 to 4.05).

Interval from diagnosis of selected conditions to study exit

For conditions identified in the multivariate model as being associated with CKD risk, there were significant differences between cats with and without CKD in the median interval between diagnosis of the condition and study exit. The median interval between cystitis diagnosis and study exit was significantly (P = 0.01) shorter for cats with (500 days) than cats without CKD (561 days). The interval between diagnosis of cystitis and CKD was 0 days (ie, diagnosed concurrently) for 190 of 743 (25.6%) CKD patients diagnosed as having cystitis. The median interval between diagnosis of diabetes mellitus and study exit was significantly (P < 0.001) greater for cats with (616 days) than cats without CKD (400 days). The median interval between diagnosis of hepatic lipidosis and study exit was not calculated, as only 2 CKD patients had a diagnosis of hepatic lipidosis.

Discussion

In the present study, we found that cats with PD had a greater risk for CKD than did cats without PD, with the highest risk associated with stages 3 and 4 PD. These results were similar to findings from cohort studies5,15 involving canine or human patients. Periodontal disease is a common but preventable condition in cats. Published population-based estimates of feline PD range from 3% to 13.9%, with prevalence increasing with increasing age.31,32 The estimated overall prevalence of PD (56,514/836,275 [7%]) among feline patients that had ≥ 3 visits at Banfield hospitals during the study period was consistent with these estimates, although our estimate was low, as it included only cats with PD that was diagnosed and staged. When both staged and unstaged diagnoses of PD were considered in a 2015 report,33 an estimated 21% of cats seen at Banfield hospitals nationwide had a diagnosis of PD.

Age was included in the final multivariate model and was strongly associated with risk of CKD, consistent with other reports indicating that CKD is primarily a disease of older cats.1,3,4,34 The median age of cats subjectively increased with increasing stage of PD. In addition, cats that developed CKD were significantly older than cats without CKD (median age, 11.2 vs 6.5 years, respectively). However, inclusion of age in the final model allowed for the interpretation that PD was a risk factor for CKD, independent of age.

Having undergone anesthesia within the year before study exit was associated with a significant increase in the risk of CKD. This was similar to the finding from an earlier case-control study2 to identify risk factors for CKD among cats. Anesthesia has been theorized to be a potential risk factor for acute kidney injury because of a risk of renal hypoperfusion during and after surgical procedures,35 although the routine administration of crystalloid fluids and continuous monitoring during anesthetic procedures at Banfield hospitals should have helped to mitigate this risk. Additional studies are needed to evaluate outcomes following general anesthesia in cats.

Of 55,999 cats in the present study that underwent general anesthesia within the year before study exit, 39,014 (69.7%) also had a dental cleaning during the same period. Dental procedures aimed at managing PD are among the more common reasons for anesthesia in older cats,2 so confounding among age, anesthesia, and history of dental cleaning was to be expected. In the present study, the median age of cats that had a dental cleaning was slightly, but significantly, higher than that for cats with no history of dental cleaning (7.3 vs 7.1 years, respectively). In addition, cats in the present study at all stages of PD were more likely to have had a dental cleaning, compared with cats without PD. Another consideration is that cats brought in for procedures such as dental cleanings were more likely to receive regular veterinary care and preanesthetic hematologic analysis, and therefore, CKD would be more likely to be detected (ie, detection bias would be expected). This was likely to have played an important role in the present study, as most (40,856/56,414 [72.4%]) cats that had a diagnosis of PD had been enrolled in a wellness plan at some point during the study period and thus might have been more likely to receive regular preventive care than cats that had not been enrolled in such a program. Accordingly, number of veterinary visits and history of anesthesia were included in the model to adjust for these systematic differences in the study population.

The discrepancies between groups of patients that were more likely to receive regular wellness care versus those seen on a more limited basis pose a challenge to population-based research in this setting that can be described as a wellness paradox. Pets that receive routine preventive services such as dental cleanings and vaccinations can appear to be proportionately more likely to develop adverse health conditions as a result of the increased medical scrutiny that they receive by virtue of their wellness plan coverage. For instance, in a survival analysis study7 of cats, a client-reported history of frequent or regular vaccination was found to be associated with increased risk of CKD, although no data were collected on actual visit frequency. It was speculated that the development of anti–renal tissue antibodies following vaccination might play a role in this association, but it was also possible that vaccination in that study7 was an indicator of the amount of medical attention that predisposed detection of CKD in cats. Clients that invest in a wellness package for their pet may be more likely to bring their pet to the veterinarian for routine preventive care as well as in the event of suspected illness, because unlimited visits are included as part of the wellness package. Another potential explanation is that clients whose pets are identified as having a disease may enroll their pet in a wellness plan to reduce the overall costs associated with long-term treatment and monitoring. In a study by O'Neill et al,6 it was found that having pet insurance was associated with increased risk of CKD in dogs, which was ascribed to better detection resulting from more frequent access to veterinary care and an increased tendency among ensured pets to undergo medical procedures such as blood tests.6 Although the wellness plan available at Banfield hospitals is not equivalent to pet insurance, it is a package of prepaid wellness care services and it is reasonable to assume that clients who purchase such plans will access veterinary care more frequently. Findings in the present study that supported this hypothesis included a higher median number of veterinary visits during the study period for cats with a CKD diagnosis, compared with that for cats without a CKD diagnosis (7 vs 5 visits, respectively), as well as a longer median follow-up period (873 vs 550 days, respectively).

Results from the present study suggested that certain breeds of cats (Siamese, Himalayan, and Abyssinian) had a higher risk of CKD, compared with the risk for all cats of other breeds in the study. Although most studies attempting to look at specific cat breeds are limited by the relatively small numbers of most breeds, our study sample included 19,024 purebred cats, and this provided an advantage when evaluating the most common cat breeds individually. However, it should be noted that despite the large overall number of purebred cats, most individual breeds were represented by relatively small numbers. Our results were consistent with the findings of a case-control study2 to evaluate risk factors for CKD in cats, in which cats with CKD were significantly less likely to be domestic shorthair cats than to be of other breeds. A prevalence study4 of cats with renal failure examined at veterinary colleges between 1980 and 1990 found that renal failure was recognized more than twice as often in Maine Coon Cat, Abyssinian, Siamese, Russian Blue, and Burmese breeds, compared with the mean for the overall patient population. In a retrospective case review of cats with CKD, Siamese, Persian, and Abyssinian breeds were overrepresented, compared with the study hospital population.29 A seroprevalence study28 among 106 healthy Birman cats found an unexpectedly high prevalence of azotemia, raising concerns about the need for additional testing and monitoring in cats of this breed. These collective findings indicate that consideration of cat breed could help to facilitate early diagnosis of CKD. It is also possible that purebred cats receive more frequent veterinary care than do mixed-breed cats, and this could result in increased detection of CKD. Additional research is needed to explore possible reasons for the apparent increase in risk for CKD among certain cat breeds.

Overall, cats with PD in the present study were more likely to have a diagnosis of most of the other conditions evaluated than were cats without PD; this was likely a reflection of the greater medical scrutiny that cats with PD underwent, especially given that they were more likely to be enrolled in wellness plans during the study period. In addition, the median age of cats with PD appears to increase with stage of PD, which would also predispose them to certain age-related conditions, such as hyperthyroidism, diabetes mellitus, and hypertension.25,36 After controlling for other covariates, cystitis was the only condition other than PD that was associated with a significant increase in the risk of CKD in the multivariate model. Cystitis, particularly due to bacterial urinary tract infection, is common in cats with CKD,1 with the prevalence in 1 study37 estimated at 22%. No data were extracted to determine the validity of any diagnosis other than CKD in the present study, and the manner in which individual veterinarians diagnosed cystitis and the underlying cause in each situation was unknown. Increased risk of cystitis in cats with CKD was previously identified in a case-control study by Greene et al2 in which a similar population was used, although, given the study design, it was not possible to evaluate timing of the diagnosis of cystitis relative to CKD. In the present study, the other health conditions hypothesized to be associated with CKD or PD could have been diagnosed at any point during the cat's lifetime (up to study exit). The median interval between cystitis diagnosis and study exit was significantly shorter for cats with (500 days) than cats without CKD (561 days). This raises the possibility that subclinical CKD existed in some cats and that this increased the likelihood of a cystitis diagnosis. In people, it is estimated that 75% of CKD cases are undiagnosed.9 One hundred ninety of 743 (25.6%) cystitis diagnoses in cats with a diagnosis of CKD in the present study were made concurrently with the initial CKD diagnosis. It was also possible, although less likely, that a history of urinary tract infections could have predisposed some cats to CKD.38

In the present study, diabetes mellitus was associated with a decreased risk of CKD. This finding was in contrast to results of human studies39,40 that indicate diabetic nephropathy is an important cause of end-stage renal disease. Diabetic nephropathy occurs infrequently in small animals.41 Our results were similar to those of a case-control study2 that used data from the Banfield patient population and found that cats with CKD were less likely to have a previous diagnosis of diabetes mellitus than were cats without CKD. The authors hypothesized that osmotic diuresis associated with glucosuria may promote phosphaturia, which could have a protective effect by limiting phosphate retention and its renal consequences. It was also possible that cats with diabetes mellitus were less likely to have survived long enough for CKD to be diagnosed (ie, if cats diagnosed with diabetes mellitus were more likely to be euthanized or otherwise lost to follow-up than cats without diabetes mellitus). In the present study, the median interval between diagnosis of diabetes mellitus and study exit date was significantly greater for cats with (616 days) than cats without (400 days) CKD.

The finding that hepatic lipidosis was associated with a decreased risk of CKD in the present study was unexpected. Several human studies42–44 have identified hepatic lipidosis as a risk factor for CKD. It should be noted that the number of cats with a diagnosis of hepatic lipidosis in our study was relatively small (836), and only 2 of these cats had a diagnosis of CKD, which was reflected in the large CI for the HR estimate. Hepatic lipidosis typically develops in cats with an overweight body condition when a primary disease process causing anorexia sets in.45 Cats with a diagnosis of CKD in the present study were typically older and generally weighed less than cats without CKD, which might have accounted in part for the lack of a positive association between the 2 conditions.

There were some limitations in the present study that should be noted. The diagnosis of other health conditions, such as cystitis and hepatic lipidosis, was determined on the basis of use of diagnostic codes rather than searching the records for specific diagnostic criteria. The criteria for defining cases of CKD in the present study included a record of the diagnosis together with plasma or serum creatinine concentration > 1.6 mg/dL and USG < 1.035. The 2016 International Renal Interest Society guidelines46 define mild renal azotemia in cats as blood creatinine concentrations of 1.6 to 2.8 mg/dL, which overlap the upper limit of the reference range for many laboratories. The specificity of our study criteria for CKD may thus have been somewhat low when applied to cats that had circulating creatinine concentrations at the lower end of the range. The 2016 International Renal Interest Society guidelines46 also include recommendations for the use of other measures such as proteinuria and blood pressure for substaging and the use of blood concentrations of symmetric dimethylarginine as a more sensitive biomarker of renal function than blood creatinine concentrations; however, most of these data were not available for the present study. Another consideration is that the eligibility criteria for the study were somewhat stringent, such as the requirements for a minimum number of visits, minimum follow-up time, and staging of PD. These requirements may limit the generalizability of our findings because not all cats receive this level of medical scrutiny. In addition, although cats that did not have any component of PD diagnosed were eligible for inclusion in the unexposed cohort, it is possible that some cats with undiagnosed PD were included in this cohort. It should be noted that the diagnosis of PD was made by veterinarians in any of several hospitals, and these individuals may not have necessarily relied upon radiographic evaluation to make the diagnosis. Although variability in diagnostic practices likely existed, we believe that it would have had a minimal impact on our overall findings. Lastly, previous studies47,48 have found associations between dietary factors and CKD, but information on the type of diet fed during the study period was lacking for many cats in the present study, and this precluded the ability to adjust for any potential effect of diet on the risk of CKD.

On the basis of the results of this and other studies, it appears likely that breed, in addition to age and PD, may contribute to the risk of CKD. The potential for genetic predisposition for CKD among purebred cats should be explored. Our findings supported the importance of preventing the development of PD in cats. Client education, particularly for owners of young pets, can help establish the habit of at-home dental care early on and emphasize the importance of routine veterinary examinations and dental cleanings in preventing and mitigating the effects of PD as cats age. By focusing attention on maintenance of good oral health and wellness care, veterinarians play an important role in reducing the risk for CKD in cats.

Acknowledgments

The authors declare that there were no conflicts of interest.

The authors thank Kirk Breuninger and JoAnn Morrison for their helpful suggestions and Carina Salt (WALTHAM Centre for Pet Nutrition, Mars Petcare) for statistical advice.

ABBREVIATIONS

CI

Confidence interval

CKD

Chronic azotemic kidney disease

FVRCP

Feline viral rhinotracheitis, calicivirus, and panleukopenia virus

HR

Hazard ratio

PD

Periodontal disease

USG

Urine specific gravity

Footnotes

a.

PetWare, Banfield Pet Hospital, Vancouver, Wash.

b.

SAS, version 9.4 for Windows, SAS Institute Inc, Cary, NC.

References

  • 1. Bartges JW. Chronic kidney disease in dogs and cats. Vet Clin North Am Small Anim Pract 2012;42:669692.

  • 2. Greene JP, Lefebvre SL, Wang M, et al. Risk factors associated with the development of chronic kidney disease in cats evaluated at primary care veterinary hospitals. J Am Vet Med Assoc 2014;244:320327.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3. DiBartola SP, Rutgers HC, Zack PM, et al. Clinicopathologic findings associated with chronic renal disease in cats: 74 cases (1973–1984). J Am Vet Med Assoc 1987;190:11961202.

    • Search Google Scholar
    • Export Citation
  • 4. Lulich J. Feline renal failure: questions, answers, questions. Compend Contin Educ Pract Vet 1992;14:127153.

  • 5. Glickman LT, Glickman NW, Moore GE, et al. Association between chronic azotemic kidney disease and the severity of periodontal disease in dogs. Prev Vet Med 2011;99:193200.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. O'Neill DG, Elliott J, Church DB, et al. Chronic kidney disease in dogs in UK veterinary practices: prevalence, risk factors, and survival. J Vet Intern Med 2013;27:814821.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7. Finch NC, Syme HM, Elliott J. Risk factors for development of chronic kidney disease in cats. J Vet Intern Med 2016;30:602610.

  • 8. Kshirsagar AV, Moss KL, Elter JR, et al. Periodontal disease is associated with renal insufficiency in the Atherosclerosis Risk In Communities (ARIC) study. Am J Kidney Dis 2005;45:650657.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. Fisher MA, Taylor GW. A prediction model for chronic kidney disease includes periodontal disease. J Periodontol 2009;80:1623.

  • 10. Fisher MA, Taylor GW, Shelton BJ, et al. Periodontal disease and other nontraditional risk factors for CKD. Am J Kidney Dis 2008;51:4552.

  • 11. Grubbs V, Plantinga LC, Crews DC, et al. Vulnerable populations and the association between periodontal and chronic kidney disease. Clin J Am Soc Nephrol 2011;6:711717.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12. Ioannidou E, Hall Y, Swede H, et al. Periodontitis associated with chronic kidney disease among Mexican Americans. J Public Health Dent 2013;73:112119.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13. Sharma P, Dietrich T, Sidhu A, et al. The periodontal health component of the Renal Impairment In Secondary Care (RIISC) cohort study: a description of the rationale, methodology and initial baseline results. J Clin Periodontol 2014;41:653661.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14. Iwasaki M, Taylor GW, Nesse W, et al. Periodontal disease and decreased kidney function in Japanese elderly. Am J Kidney Dis 2012;59:202209.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Nadeem M, Stephen L, Schubert C, et al. Association between periodontitis and systemic inflammation in patients with end-stage renal disease. SADJ 2009;64:470473.

    • Search Google Scholar
    • Export Citation
  • 16. Ioannidou E, Swede H, Dongari-Bagtzoglou A. Periodontitis predicts elevated C-reactive protein levels in chronic kidney disease. J Dent Res 2011;90:14111415.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17. Pradeep AR, Kathariya R, Arjun Raju P, et al. Risk factors for chronic kidney diseases may include periodontal diseases, as estimated by the correlations of plasma pentraxin-3 levels: a case-control study. Int Urol Nephrol 2012;44:829839.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18. Rawlinson JE, Goldstein RE, Reiter AM, et al. Association of periodontal disease with systemic health indices in dogs and the systemic response to treatment of periodontal disease. J Am Vet Med Assoc 2011;238:601609.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19. Scannapieco FA, Panesar M. Periodontitis and chronic kidney disease. J Periodontol 2008;79:16171619.

  • 20. Ismail G, Dumitriu HT, Dumitriu AS, et al. Periodontal disease: a covert source of inflammation in chronic kidney disease patients. Int J Nephrol 2013;2013:515796.

    • Search Google Scholar
    • Export Citation
  • 21. Ioannidou E, Swede H. Disparities in periodontitis prevalence among chronic kidney disease patients. J Dent Res 2011;90:730734.

  • 22. Paepe D, Daminet S. Feline CKD: diagnosis, staging and screening—what is recommended? J Feline Med Surg 2013;15:1527.

  • 23. Reynolds BS, Lefebvre HP. Feline CKD: pathophysiology and risk factors—what do we know? J Feline Med Surg 2013;15(suppl 1):314.

  • 24. Hsu V, Grant DC, Zajac AM, et al. Prevalence of IgG antibodies to Encephalitozoon cuniculi and Toxoplasma gondii in cats with and without chronic kidney disease from Virginia. Vet Parasitol 2011;176:2326.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25. Vaske HH, Schermerhorn T, Grauer GF. Effects of feline hyperthyroidism on kidney function: a review. J Feline Med Surg 2016;18:5559.

  • 26. Bartges JW, Willis AM, Polzin DJ. Hypertension and renal disease. Vet Clin North Am Small Anim Pract 1996;26:13311345.

  • 27. White JD, Malik R, Norris JM, et al. Association between naturally occurring chronic kidney disease and feline immunodeficiency virus infection status in cats. J Am Vet Med Assoc 2010;236:424429.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28. Gunn-Moore DA, Dodkin SJ, Sparkes AH. An unexpectedly high prevalence of azotaemia in Birman cats (lett). J Feline Med Surg 2002;4:165166.

    • Search Google Scholar
    • Export Citation
  • 29. Boyd LM, Langston C, Thompson K, et al. Survival in cats with naturally occurring chronic kidney disease (2000–2002). J Vet Intern Med 2008;22:11111117.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30. Lin DY, Wei LJ, Ying Z. Checking the Cox model with cumulative sums of martingale-based residuals. Biometrika 1993;80:557572.

  • 31. O'Neill DG, Church DB, McGreevy PD, et al. Prevalence of disorders recorded in cats attending primary-care veterinary practices in England. Vet J 2014;202:286291.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32. Lund EM, Armstrong PJ, Kirk CA, et al. Health status and population characteristics of dogs and cats examined at private veterinary practices in the United States. J Am Vet Med Assoc 1999;214:13361341.

    • Search Google Scholar
    • Export Citation
  • 33. Banfield Pet Hospital. State of pet health 2016 report. Available at: www.banfield.com/Banfield/media/PDF/Downloads/soph/Banfield-State-of-Pet-Health-Report-2016.pdf. Accessed May 2, 2016.

    • Search Google Scholar
    • Export Citation
  • 34. Elliott J, Barber PJ. Feline chronic renal failure: clinical findings in 80 cases diagnosed between 1992 and 1995. J Small Anim Pract 1998;39:7885.

    • Search Google Scholar
    • Export Citation
  • 35. Mitchell SK, Toal RL, Daniel GB, et al. Evaluation of renal hemodynamics in awake and isoflurane-anesthetized cats with pulsed-wave Doppler and quantitative renal scintigraphy. Vet Radiol Ultrasound 1998;39:451458.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36. Pittari J, Rodan I, Beekman G, et al. American Association of Feline Practitioners. Senior care guidelines. J Feline Med Surg 2009;11:763778.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37. Mayer-Roenne B, Goldstein RE, Erb HN. Urinary tract infections in cats with hyperthyroidism, diabetes mellitus and chronic kidney disease. J Feline Med Surg 2007;9:124132.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38. Huland H, Buchardt P, Kollermann M, et al. Vesicoureteral reflux in end stage renal disease. J Urol 1979;121:1012.

  • 39. Debnath S, Thameem F, Alves T, et al. Diabetic nephropathy among Mexican Americans. Clin Nephrol 2012;77:332344.

  • 40. Navarro-González JF, Mora-Fernandez C. The role of inflammatory cytokines in diabetic nephropathy. J Am Soc Nephrol 2008;19:433442.

  • 41. Muñana KR. Long-term complications of diabetes mellitus, part I: retinopathy, nephropathy, neuropathy. Vet Clin North Am Small Anim Pract 1995;25:715730.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 42. Musso G, Cassader M, Cohney S, et al. Emerging liver-kidney interactions in nonalcoholic fatty liver disease. Trends Mol Med 2015;21:645662.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 43. Byrne CD, Targher G. NAFLD: a multisystem disease. J Hepatol 2015;62:S47S64.

  • 44. Bang KB, Cho YK. Comorbidities and metabolic derangement of NAFLD. J Lifestyle Med 2015;5:713.

  • 45. Center SA. Feline hepatic lipidosis. In: Aielo SE, Moses MA, eds. The Merck veterinary manual. Kenilworth, NJ: Merck Sharp & Dohme Corp, 2015. Available at: www.merckvetmanual.com/mvm/digestive_system/hepatic_disease_in_small_animals/feline_hepatic_lipidosis.html. Accessed May 2, 2016.

    • Search Google Scholar
    • Export Citation
  • 46. International Renal Interest Society. IRIS staging of CKD (modified 2015). Available at: www.iris-kidney.com/pdf/staging-of-ckd.pdf#page=7. Accessed May 2, 2016.

    • Search Google Scholar
    • Export Citation
  • 47. Hughes KL, Slater MR, Geller S, et al. Diet and lifestyle variables as risk factors for chronic renal failure in pet cats. Prev Vet Med 2002;55:115.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 48. DiBartola SP, Buffington CA, Chew DJ, et al. Development of chronic renal disease in cats fed a commercial diet. J Am Vet Med Assoc 1993;202:744751.

    • Search Google Scholar
    • Export Citation
All Time Past Year Past 30 Days
Abstract Views 636 0 0
Full Text Views 2038 1343 58
PDF Downloads 1111 495 52
Advertisement