Evaluation of laboratory findings indicating pancreatitis in healthy lean, obese, and diabetic cats

Abstract

Background: Obesity is a risk factor for diabetes mellitus, which commonly coexists with pancreatitis in cats. However, obesity has not previously been associated with pancreatitis in cats.

Objectives: To evaluate factors affecting serum concentrations of pancreatic lipase immunoreactivity (fPLI), trypsin-like immunoreactivity (fTLI), cobalamin and folate in clinically healthy lean, overweight and obese, or diabetic cats.

Animals: Seventy-nine client-owned cats (27 healthy lean [LN, BCS 4-5/9], 30 healthy overweight and obese [OW, BCS 7-9/9], and 22 diabetic [DM]) were included.

Methods: Cross-sectional study. The cats underwent physical examination, and blood tests. Linear regression models compared differences in fPLI, fTLI, cobalamin, and folate concentrations. Fisher's exact test assessed the proportions of cats with fPLI and fTLI indicative of pancreatitis, and hypocobalaminemia. A random forest algorithm identified explanatory variables for cats having fPLI levels indicative of pancreatitis.

Results: No LN cats, while 6/30 (20%) of OW and 10/22 (45%) of DM cats had fPLI concentrations indicative of pancreatitis. Body condition score (P = .02) and body weight (P = .002) were positively associated with fPLI levels in LN and OW cats. Higher fPLI, and lower cobalamin concentrations were associated with higher age across groups.

Conclusions and clinical importance: Body condition score and body weight were associated with higher fPLI levels in nondiabetic cats. A larger proportion of OW and DM cats had fPLI concentrations indicative of pancreatitis compared to LN cats. Whether this indicates subclinical pancreatitis remains to be determined. Hypocobalaminemia was less frequent in OW compared to DM cats.

Keywords: cobalamin; feline; folate; pancreatic lipase immunoreactivity; trypsin‐like immunoreactivity.

FIGURE 1

Boxplot showing feline pancreatic lipase immunoreactivity (fPLI) serum concentrations in healthy lean (LN, n = 27), overweight and obese (OW, n = 30), and diabetic (DM, n = 22) cats. Each dot represents the result of an individual cat. The gray dotted line represent the cut‐off for fPLI indicative of pancreatitis (fPLI >5.4 μg/L) set by Idexx Laboratories (Vet Med labor GmbH, Ludwigsburg, Germany).

FIGURE 2

Boxplot showing serum cobalamin concentrations in lean (LN, n = 27), overweight and obese (OW, n = 30), and diabetic (DM, n = 21) cats. Each dot represents the result of an individual cat. Gray dotted lines represent the reference interval (290‐1500 ng/L) set by Idexx Laboratories (Vet Med labor GmbH, Ludwigsburg, Germany).

FIGURE 3

Random forest model, the 10 most important predictors of the final model generated to predict factors differing between cats with a normal serum feline pancreatic lipase immunoreactivity (fPLI; <5.4 μg/L) compared to those with fPLI (≥5.4 μg/L) indicative of pancreatitis. The algorithm was trained using 10‐fold cross‐validation with randomly selected 80% of the data with an fPLI result, including age, group (LN, OW, and DM), BCS, body weight, CBC and biochemistry variables, thyroxine, SAA and urine analysis data. The model was then validated using the remaining 20% of the data. LN, healthy lean; OW, healthy overweight and obese; DM, diabetic; BCS, body condition score; SAA, serum amyloid A.