Development of a requirement for exogenous insulin treatment in dogs with hyperglycemia

Abstract

Background

It has been suggested that overt diabetes mellitus in dogs be defined based on a persistent fasting blood glucose concentration (BGC) >144 mg/dL.

Objective

Determine the number of dogs with randomly identified hyperglycemia without insulin‐treated diabetes mellitus (ITDM) that later develop a need for exogenous insulin treatment.

Animals

A total of 1318 dogs examined at a university teaching hospital without ITDM and with randomly identified hyperglycemia.

Methods

Retrospective longitudinal study. Hyperglycemia was defined as randomly identified BGC above >112 mg/dL, moderate hyperglycemia as BGC >144 mg/dL but <200 mg/dL and pronounced hyperglycemia as BGC ≥200 mg/dL. Dogs were defined as having ITDM if they were treated with insulin. Follow‐up was attempted 7 to 12 years after hyperglycemia was documented to determine if over time dogs developed a need for exogenous insulin treatment.

Results

Twenty‐nine of 824 dogs (3.5%) with hyperglycemia and follow‐up information developed ITDM, including 3/824 dogs (0.4%) with moderate hyperglycemia, and 2/824 dogs (0.2%) with pronounced hyperglycemia. Most dogs with hyperglycemia that developed ITDM (24/29, 83%) had BGC ≤144 mg/dL. Among dogs that eventually developed a need for exogenous insulin treatment, no association was found between the degree of hyperglycemia and the time interval between documentation of hyperglycemia and diagnosis of ITDM. Logistic regression determined that BGC is not significantly associated with ITDM.

Conclusions and Clinical Importance

Most dogs with randomly identified hyperglycemia did not develop a need for exogenous insulin treatment. Other criteria could be required to augment the definition of overt DM in non‐insulin treated dogs.

Abbreviations

BGC

blood glucose concentration

DM

diabetes mellitus

ITDM

insulin treated diabetes mellitus

1. INTRODUCTION

The definition of diabetes mellitus (DM) is less clearly defined in dogs as compared with humans. ¹ It has been suggested that overt DM in dogs be defined based on a persistent measurement of fasting blood glucose concentration (BGC) >144 mg/dL. ¹ This suggestion was proposed with the intent of prompting an exploration of this definition so that it can be refined, confirmed, or refuted. ¹

Another more recent personal view advocates that DM in dogs be defined based on a fasted or unfasted BGC ≥200 mg/dL if clinical signs suggestive of DM are present and cannot be explained by another disease process, or if the dog is experiencing a hyperglycemic crisis such as diabetic ketoacidosis. ² This definition stipulates that, if there is ambiguity regarding clinical signs, BGC measurement should be repeated, or additional testing such as a urinalysis or measurement of glycosylated proteins be performed. ² In dogs with a BGC ≤200 mg/dL but >126 mg/dL, with or without a hyperglycemic crisis or clinical signs suggestive of DM, DM can be confirmed by documenting persistent fasting hyperglycemia for >24 hours, or if glycosylated protein concentrations are increased. ²

These suggestions have been based, in part, on the definitions of DM and prediabetes in humans. ³ In humans, a fasting BGC of 100 to 125 mg/dL is consistent with prediabetes, and a fasting BGC >125 mg/dL is consistent with DM. ³ Fasting is defined as not having anything to eat for at least 8 hours. ³ A BGC of 140 to 199 mg/dL measured 2 hours after a glucose tolerance test, which entails drinking a high glucose concentration liquid, is consistent with prediabetes, whereas a BGC >199 mg/dL measured as part of a glucose tolerance test is consistent with DM. ³ Additionally, a randomly identified BGC concentration ≥200 mg/dL measured at any time regardless of fasting is consistent with a diagnosis of DM in humans. ³

The importance of finding a definition for DM in dogs before exogenous insulin treatment is required is that intervention before insulin treatment is required could delay or prevent the need for insulin treatment, and potentially could reverse the pathophysiology of DM. Another reason for identifying dogs with DM before they develop a need for exogenous insulin treatment is that research in these dogs could improve our understanding of the etiology of DM in dogs. It is possible that current research into the pathophysiology of DM in dogs has been hampered by studying only dogs with end‐stage, insulin‐requiring disease. Our goal was therefore to determine the number of dogs with incidental or randomly identified hyperglycemia and without insulin‐treated diabetes mellitus (ITDM) that ultimately developed a need for exogenous insulin treatment consistent with definitive DM.

2. METHODS

A university teaching hospital electronic medical record search of all BGC measured in dogs by an automated biochemical analyzer (Vitros 4600 Chemistry System, Ortho Clinical Diagnostics, Rochester, New York) between January 1, 2010, and December 31, 2012, was performed to identify dogs with hyperglycemia and without ITDM. The glucose analyte in the automated biochemical analyzer is measured by the hexokinase method and is calibrated according to manufacturer recommendations with the Vitros calibrator with every new batch of reagents, after any required maintenance, and at least every 6 months. The search included all dogs examined at the university teaching hospital during the study time frame, including dogs examined in the emergency room. Old records were intentionally selected to allow ample time for follow‐up and for dogs with randomly identified hyperglycemia to develop ITDM. Hyperglycemia was defined as a randomly identified BGC above the reference range (>112 mg/dL), moderate hyperglycemia was defined as a randomly identified BGC >144 mg/dL but <200 mg/dL, and pronounced hyperglycemia was defined as a randomly identified BGC ≥200 mg/dL when measured using an automated biochemistry analyzer. Dogs were defined as having persistent hyperglycemia if a follow‐up BGC performed at any time after the initial BGC measurement was >112 mg/dL. Dogs were defined as having ITDM if they were treated with insulin or if they were euthanized or died at the time of a diabetic crisis such as diabetic ketoacidosis, before insulin treatment was administered. Dogs with hyperglycemia were excluded from the study if they were erroneously coded as non‐diabetic when they actually had ITDM, and if they had documented hyperglycemia after a dextrose bolus or after toxic chocolate ingestion. Dogs with ITDM were excluded because our aim was to determine the number of dogs with hyperglycemia and without ITDM that eventually develop a need for insulin treatment.

Follow‐up was attempted with all owners of dogs with randomly identified hyperglycemia to determine if over time these dogs developed ITDM. Phone and email follow‐up were attempted in the summer of 2019 and again through 2022. When medical records indicated a diagnosis of ITDM on a specific month with no mention of the day, the date of ITDM diagnosis was entered as the first of that month. When a date of ITDM diagnosis in the medical record was limited to the year, the date of ITDM diagnosis was entered as the first of January of that year. The same criteria for dating the ITDM diagnosis were used if owner recall was imprecise, and no medical records were available to verify the date of ITDM onset.

For dogs that owners reported did not develop ITDM during the 7 to 12 years that elapsed since hyperglycemia initially was documented, medical records were reviewed to determine if these dogs had another BGC documented after the initial randomly identified hyperglycemia. In dogs that did not develop ITDM, and in which several follow‐up BGC were documented after the time of initial hyperglycemia, the follow‐up BGC measured at the longest time interval after the date of the initial hyperglycemia is reported.

Medical records also were searched for conditions other than ITDM that have been reported to cause hyperglycemia in dogs. These conditions included current treatment with glucocorticoids, cyclosporine, or tramadol, hyperadrenocorticism, hepatocutaneous syndrome, bite wound injuries, blunt trauma, hypovolemic shock, pregnancy, and diestrus. ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ Dogs were defined as receiving these drugs or having these conditions if these medications and diagnoses were clearly recorded in the medical record. Dogs were defined as having hyperadrenocorticism if the medical record confirmed this diagnosis and they were treated with trilostane. Dogs were defined as having a suspicion for hyperadrenocorticism if there was adrenal axis testing that confirmed the diagnosis, but the dog was not treated with trilostane, or if there was a statement in the medical record that indicated a suspicion for hyperadrenocorticism and a recommendation for adrenal axis testing that was not pursued.

2.1. Statistical methods

Categorical variables are reported as counts and percentages. Continuous variables were not normally distributed as determined visually and by skewness and kurtosis tests and are therefore reported as median (range). The Wilcoxon‐Mann‐Whitney test was used to compare continuous variables in 2 independent samples, and the Spearman correlation test was utilized to determine if a correlation existed between continuous variables. Chi‐squared or Fisher's exact tests were employed to determine the relationship between 2 categorical variables, depending on whether the frequency of observations per cell was 5 or fewer. Simple logistic regression was performed to determine if the binary outcome of the presence or absence of ITDM was significantly associated with BGC. A P value <.05 was considered significant for all tests. All statistical evaluations were performed using a statistical software package (Stata 14.0 for Mac, Stata Corporation, College Station, Texas).

3. RESULTS

Between January 1, 2010, and December 31, 2012, 10 972 BGC from 7615 dogs were measured using an automated biochemical analyzer. Hyperglycemia was documented in 1503 of 10 972 (13.7%) of the samples obtained from 1338 of 7615 (17.6%) of the dogs in which BGC was measured. Ten dogs, accounting for 10 BGC, were excluded because they had confirmed ITDM at the time of BGC measurement and were erroneously coded as dogs without ITDM. Ten additional dogs, also accounting for 10 BGC, were deleted because BGC was measured after a documented dextrose bolus (8 dogs), toxic chocolate ingestion (1 dog), or a suspected dextrose bolus that was not documented in the medical record (1 dog). This dog that was examined for seizures had a BGC of 519 mg/dL at that time. Repeat BGC measurement in this dog 595 days later was 102 mg/dL. The remaining 1483 of 10 972 (13.5%) samples with hyperglycemia from 1318 of 7615 (17.3%) dogs were included in the analysis. Moderate hyperglycemia in which BGC was >144 mg/dL but <200 mg/dL was identified in 195 of 1318 (15%) of dogs. Pronounced hyperglycemia in which BGC was ≥200 mg/dL was documented in 58 of 1318 (4%) dogs (Figure 1).

FIGURE 1.

The number and description of blood glucose samples and dogs screened and included in the study. BGC, blood glucose concentration.

Follow‐up, after the time of hyperglycemia, was attempted with all owners of 1318 dogs that had randomly identified hyperglycemia to determine if dogs with hyperglycemia went on to develop ITDM during the 7 to 12 years that elapsed since hyperglycemia was documented. Owners of 824 dogs (62.5%) were successfully contacted to determine if and when the dog went on to develop ITDM. Owners of 349 of 1318 (26.5%) dogs did not reply after several outreach attempts, and no adequate contact information was available for owners of 145 of 1318 (11%) dogs.

Twenty‐nine of 824 dogs (3.5%) with randomly identified hyperglycemia and follow‐up information developed ITDM. Only 3 of 29 dogs (10%) that developed ITDM had moderate hyperglycemia, and 2 of 29 dogs (7%) that developed ITDM had pronounced hyperglycemia. Most dogs with hyperglycemia that later developed ITDM (24/29, 83%) had a BGC ≤144 mg/dL, and most dogs with randomly identified hyperglycemia and follow‐up information (795 of 824 dogs, 96.5%) did not develop ITDM over the 7 to 12 years of follow‐up. Median BGC in all dogs with randomly identified hyperglycemia that eventually developed ITDM was 123 mg/dL (range, 113‐281 mg/dL). Median time between documentation of randomly identified hyperglycemia until diagnosis of ITDM in these 29 dogs was 1.6 years (range, 0.1‐8.2 years). Among dogs that eventually developed ITDM, no association was found between the degree of randomly identified hyperglycemia and the time interval between documentation of randomly identified hyperglycemia and the diagnosis of ITDM (P = .7). Nine‐hundred and forty‐two random BGC were available for review in the 824 dogs with follow‐up information and a known follow‐up ITDM status. Logistic regression in the subset of 942 samples with hyperglycemia and a known follow‐up ITDM status determined that random BGC was not significantly associated with ITDM (P = .6).

Among the group of 58 of 1318 dogs (4%) that had a random BGC ≥200 mg/dL, 43 of 58 dogs (74%) had follow‐up data to determine if they went on to develop ITDM. Most of these dogs (41 of 43, 95%) did not develop ITDM, whereas 2 of 43 dogs (5%) did develop ITDM. Of the 2 dogs with pronounced hyperglycemia that developed ITDM, 1 was examined for traumatic bite wounds and developed ITDM 31.7 months (2.6 years) after the initial randomly identified hyperglycemia was documented, and the other was examined for large cell lymphoma which prompted treatment with glucocorticoids and developed ITDM 1.35 months (0.1 years) after the initial randomly identified hyperglycemia was documented.

An additional BGC measurement after the time that initial randomly identified hyperglycemia was documented was reported in 193 of the 795 (24%) dogs that owners reported did not develop ITDM during the 7 to 12 years that elapsed since hyperglycemia was initially documented. The median time interval between the initial randomly identified hyperglycemia and the time of follow‐up BGC in dogs that did not develop ITDM was 2.2 years (range, 0.02‐9 years) and was reported in 176 of 193 (91%) of these dogs. Median follow‐up BGC in dogs that did not develop ITDM was 104 mg/dL (range, 54‐201 mg/dL). Fifty‐three of the 193 dogs (27%) that did not develop ITDM and had follow‐up BGC had persistent hyperglycemia with BGC >112 mg/dL and in 5 of the dogs with persistent hyperglycemia BGC was >144 mg/dL.

Age at the time of hyperglycemia was available for review in 787/795 dogs (99%) that did not develop ITDM, and in 29/29 dogs (100%) that did develop ITDM. At the time that the initial hyperglycemia was documented, the median age of 787 dogs that were confirmed on follow‐up not to have developed ITDM was 8.0 years (range, 0.1‐17 years) and the median age of 29 dogs that were confirmed on follow‐up to have developed ITDM was 8.0 years (range, 3‐16 years). No significant difference was found between the age of dogs that did and did not develop ITDM, at the time that randomly identified hyperglycemia was documented (P = .9).

Among dogs with randomly identified hyperglycemia that developed ITDM, 13/29 (45%) had a condition other than DM that could contribute to hyperglycemia. Six of these 13 dogs (46%) were treated with glucocorticoids, 5 (38%) had trilostane‐treated hyperadrenocorticism, 1 (8%) had hepatocutaneous syndrome, and 1 (8%) was examined for bite wound injuries at the time the initial randomly identified hyperglycemia was documented. Glucocorticoids were administered to 1 dog each for granulomatous meningoencephalitis, pemphigus foliaceus, immune‐mediated thrombocytopenia, atopy, degenerative intervertebral disc disease, or upper airway disease of unknown etiology evaluated at the referring veterinarian.

Among dogs with randomly identified hyperglycemia that did not develop ITDM, 400/795 dogs (50%) had a condition other than DM that could have contributed to hyperglycemia. At the time the initial randomly identified hyperglycemia was documented, 244 of these 400 dogs (61%) had been treated with glucocorticoids, 54 (13.5%) had trilostane‐treated hyperadrenocorticism, 45 (11.25%) had hypovolemic shock, 25 (6.25%) were treated with tramadol, 12 (3%) had experienced blunt trauma, 8 (2%) had a suspicion for hyperadrenocorticism, 6 (1.5%) were examined for bite wound injuries, 5 (1.25%) had received cyclosporine, and 1 (0.25%) had hepatocutaneous syndrome. The most common reasons for glucocorticoid administration were lymphosarcoma (35 of 244, 14%), degenerative intervertebral disc disease (20 of 244 dogs, 8%), mast cell tumor (17 of 244 dogs, 7%), meningoencephalitis (12 of 244 dogs, 5%), immune‐mediated anemia or thrombocytopenia (10 of 244 dogs, 4%), inflammatory bowel disease (5 of 244 dogs, 2%), and in 3 of 244 dogs each (1%), hypoadrenocorticism, pemphigus foliaceus, or other neoplasia. No significant difference was found in the frequency of conditions other than DM that could contribute to hyperglycemia in dogs that did or did not develop ITDM (P = .6). Specifically, no significant difference was found in the frequency of glucocorticoid administration in dogs that did or did not develop ITDM (P = .2).

4. DISCUSSION

Most dogs (≥95%) with randomly identified hyperglycemia and follow‐up information did not develop a need for exogenous insulin treatment over the 7 to 12 years of follow‐up. This finding held true not only for dogs with any hyperglycemia, but also for dogs with moderate and pronounced hyperglycemia. This finding emphasizes the need to introduce criteria other than randomly identified hyperglycemia to augment the definition of overt DM in dogs. It also calls into question the importance of hyperglycemia as a sole criterion for the definition of overt DM in dogs. Finally, the need to introduce preventive measures to delay or prevent the onset of ITDM in dogs with randomly identified hyperglycemia must be weighed against the evidence suggesting that most dogs with hyperglycemia do not develop a need for exogenous insulin treatment.

In humans with type 2 DM, preventive measures are aimed at decreasing the risk of progression of fatty liver and cardiovascular disease, and include lifestyle changes such as improved nutrition, physical activity, sleep hygiene, and healthy habits directed at achieving weight loss. ¹² Medical management of hypertension and dyslipidemia is also important. ¹² Although obesity likely is not associated with DM in dogs, hepatic lipidosis, hypertension, and dyslipidemia have been reported in dogs with DM. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ Treatment directed at alleviating these conditions is recommended in dogs, regardless of whether they are at risk for DM. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ Therefore, it is not crucial to identify the risk of DM to introduce preventive measures directed at treating hepatic lipidosis, hypertension, and dyslipidemia in dogs.

In humans with type 1 DM, early detection of disease risk involves genetic testing and identification of islet‐specific autoantibodies. ¹⁷ Genetically predisposed individuals are thought to encounter a triggering event that begins a process of autoimmune destruction of beta cells culminating in insulin deficiency and type 1 DM. ¹⁷ Additionally, most children with ≥2 islet‐specific autoantibodies progress to type 1 DM by the age of 20. ¹⁷ Clinical trials examining various immunotherapies directed at preventing disease progression, including anti‐CD3 monoclonal antibodies, have reported some encouraging and other unsuccessful results. ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ Given inconsistent evidence of beta‐cell autoimmunity in dogs with DM, the need for early detection of dogs with DM for the purpose of introducing similar therapeutics is questionable. ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶

Although currently available therapeutic options directed at prevention of DM progression might not be applicable to dogs at this time, there is still value in early identification of dogs with DM. ²³ Early identification of dogs with DM before there is a need for exogenous insulin treatment could facilitate the study of the pathophysiology of DM in dogs. For example, in humans at risk for type 1 DM, islet‐specific autoantibodies are detected years before the onset of clinical disease, but studies in dogs are limited to dogs with overt DM. ¹⁷ , ²¹ , ²² , ²³ It is possible that an examination of autoimmunity in dogs early in the course of the natural history of DM would allow for detection of islet‐specific autoantibodies. ²³

In addition to finding that most dogs with randomly identified hyperglycemia, random moderate hyperglycemia, and randomly identified pronounced hyperglycemia did not develop a need for exogenous insulin treatment, we also found that most dogs with randomly identified hyperglycemia that did develop ITDM had a BGC ≤144 mg/dL, that no association existed between the degree of randomly identified hyperglycemia and the time interval between documentation of hyperglycemia and the diagnosis of ITDM, and that the random BGC was not a significant predictor of ITDM. If a high random BGC alone was a useful predictor of ITDM, one would expect that with a higher BGC, the time interval between documentation of hyperglycemia and the diagnosis of ITDM would be shorter. All of these findings support the conclusion that randomly identified hyperglycemia alone might not be a useful single criterion for the definition of overt DM in dogs. However, future studies examining the association between glycosylated proteins in dogs without ITDM and the onset of ITDM could yield different results. Future prospective studies also could examine the utility of repeating BGC at fixed time intervals to determine if certain time intervals are better than others in predicting eventual development of ITDM.

Many dogs in our study had conditions other than DM that could have contributed to randomly identified hyperglycemia, and it is possible that such hyperglycemia was documented not because dogs were in the early stages of DM, but rather because they had concurrent illnesses or conditions contributing to transient hyperglycemia. The most common concurrent illnesses or conditions identified in dogs with hyperglycemia that did and did not go on to develop ITDM were glucocorticoid administration and hyperadrenocorticism. Clinical signs would likely not help distinguish glucocorticoid administration, hyperadrenocorticism, and DM because all 3 conditions have similar clinical signs. No significant difference was found in the frequency of glucocorticoid administration in dogs that did and did not ultimately develop ITDM. This finding suggests that the hyperglycemia noted in our study is not a marker for insidious DM, but perhaps a finding associated with concurrent illnesses or conditions such as glucocorticoid administration. The presence of pancreatitis was not reported in our study because conclusive evidence that naturally occurring pancreatitis causes hyperglycemia in dogs without DM is lacking. ¹ , ²⁷

The incidence of DM in dogs in the United States is not known. However, the incidence of DM in dogs in Sweden, where only approximately 7% of female dogs are neutered and 72% of dogs with DM are female, is 13 cases per 10 000 dog‐years at risk. ²⁸ The prevalence of DM in dogs has been reported to be 0.6% in the United States, 0.3% in the United Kingdom, and 0.4% in Australia. ²⁹ , ³⁰ , ³¹ Different study methodologies and population neuter status complicate comparisons of these findings, because all but our study relied on large registries, and ours is the only study in which owners were contacted for verification of the diagnosis of ITDM. However, in our study 29 of 824 dogs (3.5%) with hyperglycemia and follow‐up information developed ITDM, and this percentage is higher than the prevalence of DM in dogs reported in the general population in other studies. This observation suggests that whereas most dogs with hyperglycemia do not develop a need for exogenous insulin treatment, it is possible that the population of dogs with hyperglycemia does develop ITDM more commonly than does the population of dogs at large. Interestingly, among the population of dogs with hyperglycemia, only 3/824 dogs (0.4%) with moderate hyperglycemia and 2/824 dogs (0.2%) with pronounced hyperglycemia developed ITDM, indicating that among dogs with hyperglycemia the incidence of ITDM in dogs with moderate or pronounced hyperglycemia is similar to previously reported prevalences of DM in the general dog population. ²⁹ , ³⁰ , ³¹ A potentially important future study stemming from our findings would be comparing the incidence of ITDM in dogs with and without various degrees of randomly identified hyperglycemia in order to determine whether randomly identified hyperglycemia is indeed a risk for ITDM in dogs.

The median age of dogs at the time that randomly identified hyperglycemia was documented was 8 years, whether or not they went on to develop ITDM. This is similar to the median age of diagnosis of ITDM in United States dogs, which is 8.7 years. ³² However, some of the dogs in our study were very young at the time that randomly identified hyperglycemia was documented. Juvenile ITDM is rare, and the young dogs with randomly identified hyperglycemia were therefore unlikely to have ITDM at the time the initial hyperglycemia was documented. ³² However, follow‐up on these dogs was obtained 7 to 12 years after the time that the initial randomly identified hyperglycemia was noted. Therefore, at the time of follow‐up, when a determination was made as to whether the dog had developed ITDM, most dogs would have reached the median age of diagnosis of ITDM in dogs.

One of the limitations of our study is that medical records did not specifically note if dogs were fasted. However, standard of care for our hospital is to require all dogs, except dogs with ITDM, to be fasted by withholding food but not water, overnight before their examination. Owners are reminded of this several times before the appointment in writing and over the phone and are warned that testing might not be possible if their pet is not fasted overnight before the visit. Therefore, most BGC likely were measured in fasted dogs. However, the American Diabetes Association defines a random BGC that does not require fasting of ≥200 mg/dL as consistent with a diagnosis of DM in humans, and in our study 95% of dogs with a BGC ≥200 mg/dL did not go on to develop ITDM. ³ Therefore, tools other than those provided by the American Diabetes Association likely are needed for the definition of overt DM in dogs. Although direct comparisons between species are difficult to make, the suggestion that overt DM in dogs be defined based on a persistent fasting BGC >144 mg/dL was made based on definitions of DM in humans, with a call to veterinary colleagues to explore this definition further. ¹

Another study limitation is that not all dogs had documented persistent hyperglycemia. However, a subset of 53 dogs that did not develop ITDM did have documented persistent hyperglycemia, indicating that persistent hyperglycemia alone also is not a sufficient single criterion for the diagnosis of overt DM in dogs. Additionally, dogs were defined as having hyperadrenocorticism if they were treated with trilostane and were defined as having a suspicion for hyperadrenocorticism if adrenal axis testing confirmed the diagnosis but the dog was not treated with trilostane, or if a statement in the medical record indicated a suspicion for hyperadrenocorticism and a recommendation for adrenal axis testing that was not pursued. It is therefore possible that some dogs with a suspicion for hyperadrenocorticism did in fact have hyperadrenocorticism. Furthermore, the retrospective study design did not allow for a determination of whether stress contributed to hyperglycemia in some of the dogs. Finally, our study, which aimed to determine the number of dogs with hyperglycemia and without ITDM that develop a need for exogenous insulin treatment over time, did not determine whether mild hyperglycemia is a risk factor for ITDM in dogs. Future studies comparing the incidence of ITDM in dogs with and without mild hyperglycemia are needed to address this question.

In conclusion, most dogs with randomly identified hyperglycemia, including dogs with pronounced hyperglycemia and persistent hyperglycemia, do not develop a need for exogenous insulin treatment. Most dogs with randomly identified hyperglycemia that developed ITDM had a BGC above the normal reference range, but they did not have moderate or pronounced hyperglycemia. Indeed, BGC was not a significant predictor of ITDM. No association was found between the degree of randomly identified hyperglycemia and the time to ITDM diagnosis in the subset of dogs that did develop a need for exogenous insulin treatment. Criteria other than randomly identified hyperglycemia therefore could be required to augment the definition of overt DM in non‐insulin treated dogs.

CONFLICT OF INTEREST DECLARATION

Authors declare no conflict of interest.

OFF‐LABEL ANTIMICROBIAL DECLARATION

Authors declare no off‐label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION

Authors declare no IACUC or other approval was needed.

HUMAN ETHICS APPROVAL DECLARATION

Authors declare human ethics approval was not needed for this study.

DiNinni A, Hess RS. Development of a requirement for exogenous insulin treatment in dogs with hyperglycemia. J Vet Intern Med. 2024;38(2):980‐986. doi: 10.1111/jvim.16990