Feline diabetes mellitus is a common endocrine problem in cats but much about the disease remains unknown. Fortunately, feline diabetes is a topic of active research at various institutions around the world and as a result, significant advances have been made in the management of this disease. The gains made in diabetes management are reflected in the higher diabetes prevalence and lower fatality rate at presentation demonstrated in a recent study of diabetic cats (Prahl, et al, 2007). While an increase in the disease incidence may be partially responsible for the increase in the diabetes prevalence in the feline population over the years, it is also likely that diabetic cats live longer than in the past. The reduced case fatality rate at the cat's initial presentation for treatment probably reflects a change in owners' willingness to pursue treatment (i.e., less cats are euthanized at diagnosis) as well as improved treatment methods (fewer diabetic cats die of diabetes). This review will focus on select research advances over the past several years that have relevance for our understanding and clinical management of this important feline disorder.
Update on Epidemiology and Etiology of Feline Diabetes
Epidemiologic studies are difficult to perform but the information gained from such studies can be extraordinarily helpful for understanding the behavior of a disease within a population. It is presumed that feline diabetes, like human diabetes, has a multifactorial pathology that involves genetic and environmental factors. As such, an epidemiological approach may be an effective way to identify common risk factors and pathological features associated with feline diabetes. Results from several recent studies have added to our understanding of the epidemiology of feline diabetes. Previous estimates of diabetes prevalence in the feline population ranged from 1 in 50 to 1 in 400. New studies from Australia (Lederer, et al, 2007) and the UK (McCann, et al, 2007) established prevalences of 7.4 per 1000 cats and ~4.5 per 1000 cats, respectively, which are similar to earlier estimates. However, both the UK and Australian studies showed a clear increase in diabetes risk for cats of the Burmese breed that has not been observed in US studies. Together the epidemiologic studies indicate that the prevalence of diabetes in the overall feline population is roughly the same worldwide but substantial differences in prevalence may be observed within specific subpopulations (e.g., the Burmese breed).
Previously identified risk factors for diabetes in cats include above average body weight, male gender, and advanced age. Obesity is known to produce insulin resistance in cats and obesity has also been identified as a possible risk factor for diabetes. Previously established risk factors were also found to be associated with diabetes in the recent UK study (McCann, et al, 2007). Additional factors identified by McCann and colleagues were a history of corticosteroid use (all cats) or megestrol acetate use (male cats). Another study identified confinement indoors and physical inactivity are independent diabetes risk factors (Slingerland, et al, 2007). While the picture of feline diabetes as painted by epidemiologic studies is far from complete, important information about genetic and environmental risk factors associated with the disease has emerged in recent years. Careful epidemiologic analysis may permit subsets of feline diabetes to be defined and may uncover important information about disease etiology, suggest alternative therapies, and influence the direction of research.
Update on Early Diagnosis of Feline Diabetes
Diabetes is usually not a difficult diagnosis to make. Most cats have hyperglycemia, glucosuria, and clinical signs at the time they are identified as diabetic. However, given the similarities noted between feline diabetes and human type 2 diabetes, it is likely that cats with very early diabetes pass through a 'pre-diabetic' state characterized by hyperinsulinemia and glucose intolerance but a lack of persistent hyperglycemia and clinical signs of diabetes. At present, there is no practical way to identify cats that are at-risk for diabetes or in the earliest stages of the disease. However, the recent development and validation of an assay for feline proinsulin promises to expand the diagnostic window for feline diabetes (Kley, et al, 2008). Proinsulin is secreted along with insulin by pancreatic beta cells. In humans, proinsulin secretion is abnormal in obesity and diabetes and measurement of serum proinsulin levels is a sensitive assay of beta cell function. Using the feline proinsulin assay they developed, Hoenig and colleagues demonstrated that the ratio of proinsulin to insulin secreted in response to acute stimulation with glucose was much higher in obese cats when compared with lean cats (Kley, et al, 2008). The changes in proinsulin secretion observed in obese cats by Kley et al are also observed in lean and obese humans who go on to develop type 2 diabetes. Additional studies are needed before the feline proinsulin assay can be recommended as a diagnostic test for feline diabetes. However, measurement of proinsulin may be useful for identifying diabetic cats earlier in the evolution of the disease than is possible with currently available methods. Early identification of diabetic cats may prevent earlier veterinary intervention in the form of risk factor reduction or preventative therapy.
Update on Etiology of Feline Diabetes
Acromegaly is caused by abnormal growth hormone secretion from a pituitary tumor. Most cats with acromegaly develop glucose intolerance and diabetes from profound insulin resistance caused by growth hormone excess. Growth hormone exerts its effects through stimulation of IGF-1 production by the liver and IGF-1 levels are frequently used for making a diagnosis of acromegaly when a growth hormone assay in unavailable. The serum IGF-1 level is used to confirm the diagnosis of acromegaly when other clinical signs suggestive of the disorder are present. IGF-1 is elevated in some diabetic cats without signs of acromegaly. The serum IGF-1 concentration is altered by insulin treatment in diabetic cats but studies suggest that it cannot be used to gauge the degree of diabetic control (Berg, et al, 2007) or to distinguish cats that may have transient diabetes from those with permanent diabetes (Alt, et al, 2007b). A recent study that examined diabetic cats with high levels of IGF-1 found an unexpected high prevalence of acromegaly in the study population. Niessen et al (2007) studied a group of 184 diabetic cats for which serum IGF-1 levels were available and identified 59 cats with serum IGF-1 levels compatible with acromegaly (>1000 ng/mL). Eighteen of the 59 cats were available for complete evaluation, including brain MRI. Acromegaly was confirmed in 17 of the 18 cats that received complete evaluation. While the ramifications of these new findings regarding acromegaly and IGF-1 measurement for general diabetes management in cats are not yet fully understood, it appears that acromegaly, which has historically been considered a rare etiology of diabetes in cats, may not be as uncommon as previously thought.
Update on Diabetic Monitoring
Blood glucose monitoring in diabetic cats has traditionally been performed at the veterinary hospital under the supervision of trained veterinary staff. Within the last decade, several teams of investigators have developed methods for glucose monitoring in the home environment that include owner-generated serial blood glucose curves (e.g., Kley, et al, 2004) and the use of continuous glucose monitoring technology (e.g., Wiedmeyer, et al, 2003). These methods represent important additions to the tools used for diabetic monitoring in cats.
At-Home Glucose Monitoring
The feasibility of training cat owners to perform at-home glucose monitoring has been demonstrated (Kley, et al, 2004). Kley et al trained owners of diabetic cats to perform periodic at-home glucose curves using a commercial lance to obtain capillary blood samples, which the owner then analyzed on a portable glucometer. The owner-generated glucose curves along with periodic veterinary examinations were used to make recommendations for insulin therapy. Most cat owners successfully performed long-term home glucose monitoring and reported little or no difficulty performing the procedure beyond the initial training period. Long term monitoring (>2 years) by owners was performed in over 30% of cats. Problems with at-home monitoring included owner attrition, patient intolerance of the procedure, owner non-compliance, and technical problems (Kley, et al, 2004). Cat owners generally express a high level of satisfaction with at-home monitoring. However, successful implementation of an at-home monitoring is likely to depend on the willingness of the veterinary team to work through problems with owners. A substantial time commitment from the veterinarian and the veterinary staff should be anticipated if at-home monitoring is implemented as a routine part of diabetes management in the practice (Schermerhorn, 2005). In addition to the time and expense, other potential drawbacks of at-home monitoring include treatment errors caused by inaccurate blood glucose measurement, patient discomfort caused by frequent blood sampling, and the induction of stress or aversion behaviors in the cat. In the author's opinion, at-home monitoring should not be a routine part of diabetic monitoring in cats. At present, it seems that at-home monitoring may be beneficial for select individual cases, especially when data from in-hospital glucose curves is at odds with other clinical assessments. The use of home monitoring techniques in diabetic cats has been recently reviewed (Reusch, et al, 2006).
Continuous Glucose Monitoring
Nearly coincident with reports of the successful use of at-home serial blood glucose monitoring, the application of an electronic technology for continuous glucose monitoring to veterinary patients was described (Wiedmeyer, et al, 2003). The continuous monitoring system uses a sensor inserted in the subcutaneous space to measure the interstitial glucose concentration. Measurements of interstitial glucose are obtained every 5 minutes and stored for up to several days (Ristic et al, 2005). Interstitial glucose concentrations in normal and diabetic cats correlate well to serum glucose concentrations measured by glucometer or chemistry analyzer (Wiedmeyer, et al, 2003; Ristic et al, 2005). Wiedmeyer et al (2003) demonstrated that interstitial glucose concentrations broadly reflected the diabetic status (uncontrolled vs. controlled) determined by other clinical and laboratory methods. These promising studies suggest that diabetic cats can be effectively monitored using a 'hand-off' continuous glucose monitoring system. The clinical use of continuous blood glucose monitoring in veterinary medicine was the subject of a recent review by Wiedmeyer and DeClue (2008).
Variability in Blood Glucose Curves
Despite the widespread recommendation for the use of blood glucose curves to monitor the effectiveness of diabetic management, the technique is unreliable, inaccurate and suffers from poor repeatability. In one study, the large degree of variability between serially obtained glucose curves observed in dogs would have led to conflicting recommendations for insulin therapy in >50% of patients (Fleeman and Rand, 2003). Significant variability is also observed in diabetic cats when curves obtained at-home are compared with those obtained in-hospital (Casella, et al, 2005) or when consecutive at-home curves are compared (Alt, et al, 2007a). The variability detected in the canine and feline studies likely has numerous sources, not all of which can be adequately controlled. Sources of variability that might influence glucose curve results include analytic variability (human error introduced when obtaining or analyzing the sample), technologic variability (variability/error caused the instrumentation used to make the measurements), and individual variability (caused by inherent biologic fluctuations).
References are available upon request.