Can we predict hypoadrenocorticism in dogs with resting hypocortisolemia? A predictive model based on clinical, haematological, and biochemical variables

Background: A resting cortisol concentration (RC) higher than 2.0 μg/dL (55 nmol/L) is commonly used to rule out hypoadrenocorticism (HA). However, there is a significant overlap of RC between dogs with HA and those with other diseases. There is a need for data to help increase the suspicion of HA. Hypothesis/objectives: To create a predictive model based on clinical, haematological and biochemical variables to increase the likelihood of diagnosis of hypoadrenocorticism. Animals: Ninety-two dogs with RC <2.0 μg/dL, and an adrenocorticotropic hormone stimulation test (ACTHst) performed. Methods: Multicentric retrospective cohort study with review of medical records of client-owned dogs presented to two referral centres, between January 2018 and March 2022. Dogs were divided into two groups (HA and not HA), based on ACTHst results. Descriptive analysis was performed along with a predictive model, using univariable and multivariable logistic regression analysis. Results: Of the 92 included dogs, hypoadrenocorticism was diagnosed in 29 patients (32.2%) based on ACTHst results (HA group). Acute gastrointestinal signs, anorexia and lethargy were more prevalent in the HA group. Creatinine, BUN, ALT, and potassium were higher, and albumin, sodium and Na/K ratio were lower in the HA group. Multivariate analysis developed a robust model in which lethargy (OR 5.25), anorexia (OR 3.69), albumin (OR 0.32), and sodium (OR 0.84) concentrations allowed the prediction of HA. Conclusions and clinical importance: In dogs with resting hypocortisolemia, the combination of anorexia and lethargy, along with low sodium and albumin concentrations, should raise the suspicion of hypoadrenocorticism. The variables identified in this study may help clinicians to identify dogs with hypoadrenocorticism in daily clinical practice.