Evaluation of insulin-like growth factor-1,total thyroxine,feline pancreas-specific lipase and urinary corticoid-to-creatinine ratio in cats with diabetes mellitus in Switzerland and the Netherlands

Abstract

Objectives

The aim of the study was to evaluate circulating insulin-like growth factor-1 (IGF-1), feline pancreas-specific lipase (fPLI) and total thyroxine (TT4) concentrations and urinary corticoid-to-creatinine ratio (UCCR) as indicators for the prevalence of acromegaly, pancreatitis, hyperthyroidism and hypercortisolism in cats with diabetes mellitus.

Methods

Blood and urine samples were collected from diabetic cats treated in primary care clinics in Switzerland and the Netherlands. Standardised questionnaires and physical examination forms provided clinical information from owners and veterinarians. Laboratory testing included serum biochemistry profile analysis and measurement of circulating fructosamine, IGF-1, fPLI, and TT4 concentrations and UCCR. CT of the pituitary gland was performed using a multidetector computed tomography scanner.

Results

Blood samples were available from 215 cats and urine samples were collected at home from 117 cats. Age ranged from 2–18 years (median 12 years) and body weight from 2.7–12.3 kg (median 5.5 kg). Sixty-five percent of the cats were castrated male and 35% were female (33% spayed); 82% were domestic shorthair cats. Eighty percent of cats received a porcine insulin zinc suspension, 19.5% insulin glargine and 0.5% a human neutral protamine hagedorn insulin. Thirty-six of 202 (17.8%) cats had IGF-1 concentrations >1000 ng/ml. Serum fPLI, and TT4 concentrations and UCCR were increased in 86/196 (43.9%), 9/201 (4.5%) and 18/117 cats (15.3%), respectively. Prevalence did not differ between countries.

Conclusions

Hyperthyroidism is rare, whereas increased fPLI concentration, possibly reflecting pancreatitis, is common in diabetic cats. The high UCCR may reflect activation of the hypothalamus–pituitary–adrenal axis, which also occurs in diabetic humans. The percentage of cats with increased IGF-1 was high but lower than reported in recent studies.

Introduction

Traditionally, the classification of diabetes mellitus in cats has followed the scheme used in human medicine. Although the aetiopathogenic mechanisms may not be completely identical, the ‘human model’ provides a guide for identification and differentiation of the various forms of the disease.¹ Type 1 diabetes results from beta(β)-cell destruction, usually leading to absolute insulin deficiency. Type 2 diabetes is due to a progressive insulin secretory defect on the background of insulin resistance.²

It has been estimated that at least 80% of diabetic cats have type 2-like diabetes mellitus.^3,4 If this estimate is correct, up to 20% of diabetic cats can be assumed to have another type of diabetes. As type 1-like diabetes is extremely rare in cats, most of those 20% of cats may develop the disease as a sequel to another condition. In humans the latter category is known as ‘specific types of diabetes due to other causes’,² whereas in companion animals the term ‘secondary diabetes’ is sometimes used. Secondary diabetes may be the result of disorders of the exocrine pancreas, various endocrinopathies (eg, hypersomatotropism, hypercortisolism, hyperthyroidism) and other conditions. However, the allocation of 80% of feline diabetes cases to type 2-like diabetes and the remaining 20% to other types of diabetes is merely a clinical estimate that, to our knowledge, has not been scientifically substantiated.

Hypersomatotropism (acromegaly) has long been considered a rare disease in cats, but this belief has come under scrutiny based on the results of screening studies performed in the UK. Two studies showed that serum insulin-like growth factor-1 (IGF-1) concentration, which reflects the 24 h growth hormone secretion and is most commonly used as a screening test for hypersomatotropism, was increased in 32% and 26.1% of diabetic cats.^5,6 Pancreatitis has been recognised as a common concurrent disease in cats with diabetes mellitus, but the cause and effect relationship between the two diseases remains largely unknown. Feline pancreas-specific lipase (fPLI) concentration, an indicator of pancreatitis widely used as screening test for pancreatitis, was increased in 83% of diabetic cats in which the duration of diabetic treatment was unknown, and in 33% of cats with newly diagnosed diabetes.^7,8 Information about the prevalence of hyperthyroidism and hypercortisolism in cats with diabetes is limited. In a retrospective study carried out almost 20 years ago, 5.8% of diabetic cats tested had increased total thyroxine (TT4) concentrations and were considered hyperthyroid, and 0.9% of these cats were diagnosed with hypercortisolism.⁹ To our knowledge, systematic screening for the diseases mentioned above in a given population of diabetic cats has not been undertaken.

The aim of the study was to evaluate the circulating concentrations of IGF-1, fPLI and TT4, and determination of the urinary corticoid-to-creatinine ratio (UCCR) as surrogate markers for the prevalences of acromegaly, pancreatitis, hyperthyroidism and hypercortisolism, respectively, in cats with diabetes mellitus being managed by primary care veterinarians in two European countries. In addition, variables relating to signalment, history, complete blood counts, serum chemistry profiles and urinalysis of cats with normal and abnormal IGF-1, fPLI, TT4 and UCCR results were compared.

Materials and methods

Study design and procurement of samples

The study was designed as a prospective clinical trial and was carried out simultaneously in Switzerland and the Netherlands between March 2011 and March 2012. Veterinary clinics in the eastern part of Switzerland and in four different regions of the Netherlands were offered free evaluation of serum biochemistry profile and measurement of serum fructosamine, TT4, IGF-1, fPLI and UCCR in cats with diabetes mellitus. Participants were urged not to choose cases selectively but instead to submit samples from all diabetic cats presented to their practice during this period, regardless of the quality of glycaemic control. To be included in the study, anti-diabetic treatment (insulin and non-insulin anti-diabetic treatment, including diabetes diet) had to have been instituted for at least 4 weeks in each of the cats.

Blood samples were collected and centrifuged in the primary care veterinary clinics, and urine samples were collected by the owners at home before they went to the veterinary clinic. All serum and urine samples were stored at 4°C and were picked up within 12 h of collection by study personnel.

Standardised history taking and physical examination

Owners were asked to complete a questionnaire that provided information on signalment, duration of diabetes, treatment, general wellbeing and clinical signs potentially associated with the diseases in question (see supplementary material Appendix 1). The veterinarians were asked to provide results of the physical examination according to a standardised form (see supplementary material Appendix 2).

Analytical procedures

Serum biochemical analysis was performed using a Cobas Integra 800 (Roche Diagnostics) and standard procedures recommended by the International Federation of Clinical Chemistry. Concentrations of fructosamine, bilirubin, blood urea nitrogen, creatinine, albumin, total protein, cholesterol, triglycerides and electrolytes (sodium, potassium, calcium and phosphorus), and the activities of alkaline phosphatase (ALP), aspartate aminotransferase (AST), alanine aminotransferase (ALT) and 1,2-o-dilauryl-rac-glycero-3-glutaric acid-(6’methylresorufin) ester (DGGR) lipase were determined.¹⁰ Urinalysis included determination of urine specific gravity and urine protein-to-creatinine ratio.

TT4 concentrations were measured by chemiluminescence (Immulite 1000 Total T4; Siemens Healthcare). A TT4 ⩾45 nmol/l was considered consistent with hyperthyroidism.¹¹ IGF-1 was analysed by IGF binding protein-blocked radioimmunoassay (Mediagnost). IGF-1 concentrations >1000 ng/ml were considered consistent with acromegaly.^5,6,12

Serum fPLI was measured by a commercial ELISA (Spec fPL; IDEXX Laboratories) and concentrations >5.3 µg/l were considered consistent with suspect pancreatitis.¹³ Urinary corticoid concentrations were measured by an in-house radioimmunoassay and were related to the urinary creatinine concentrations (Jaffe kinetic method) as described by Rijnberk et al.¹⁴ The previously established reference interval for the UCCR was 8–42 × 10⁻⁶.¹⁵

Evaluation of the serum biochemistry profile and the measurement of serum fructosamine, TT4 and IGF-1 were undertaken at the Clinical Laboratory of the Vetsuisse Faculty University of Zurich. Serum fPLI was measured at the GI Laboratory, Texas A & M University, and the UCCR was measured at the Department of Clinical Sciences of Companion Animals, Utrecht University. Before analysis, samples were stored at −80°C and transported on dry ice. IGF-1, fPLI and UCCR were analysed in batches. Blood glucose concentrations were measured immediately after blood collection in the respective primary care veterinary clinic using various methods.

Computed tomography

In cats with IGF-1 concentrations >1000 ng/ml, intracranial CT was offered. At the Vetsuisse Faculty in Zurich, CT was carried out using a multidetector CT scanner (Brilliance 16-slice CT; Philips AG) with cats in sternal recumbency. Precontrast, dynamic and postcontrast CT images were obtained. Slice thickness was 1 mm with increments of 0.5 mm. A non-ionic contrast medium was used (Accupaque 350, 350 mg iodine/ml) at a dose of 2 ml/kg body weight intravenously. Pituitary attenuation was compared with brain parenchyma before contrast injection. In the dynamic study, location and symmetry of the pituitary flush were investigated for mass effect. Pituitary size was determined as pituitary height-to-brain area ratio, whereas attenuation and homogeneity of enhancement were evaluated subjectively.

At the University of Utrecht, CT images were obtained with a single-detector helical CT unit (Philips Secura; Philips NV), using 2 mm-thick transverse slices with a pitch of 1 (120 kV, 180–260 mA), both before and following intravenous administration of 2 ml/kg contrast medium (Telebrix 350, sodium and meglumine joxitalamate, containing 350 mg iodine/ml; Guerbet Nederland BV). The cats were in sternal recumbency.

Statistical analysis

A commercial software package (SAS, version 9.3; SAS Institute) was used for all statistical analyses. Differences between cat populations in Switzerland and the Netherlands were analysed using non-parametric statistical methods. Differences in the distribution of sex, breed (purebred or crossbred), types of insulin used (porcine insulin zinc suspension or insulin glargine) and cats with IGF-1 >1000 ng/ml, fPLI >5.3 µg/l, TT4 ⩾45 nmol/l or UCCR >42 × 10⁻⁶ were analysed using a Fisher’s exact test, and differences in age, body weight and insulin dosage in diabetic cats were analysed using a Kruskal–Wallis rank sum test.

One-way ANOVA was used to examine differences between cats with IGF-1 >1000 and ⩽1000 ng/ml, cats with fPLI >5.3 and ⩽5.3 µg/l, cats with TT4 ⩾45 and <45 nmol/l, and cats with UCCR >42 and ⩽42 × 10⁻⁶ for the following variables: age, body weight, duration of diabetes, insulin dose, score of polyuria and polydipsia (both with increasing severity from 1 to 10), biochemical variables (serum concentrations of glucose, fructosamine, bilirubin, urea, creatinine, albumin, total proteins, cholesterol, triglycerides, sodium, potassium, calcium, phosphate), enzyme activities (ALP, AST, ALT, DGGR lipase) and urinary variables such as specific gravity and protein-to-creatinine ratio. Because of a potential confounding effect of country (Switzerland and the Netherlands) on some variables, the country of origin was used as a blocking factor in all analyses to reduce the effect of this nuisance parameter. Normal distribution of the data was assessed using a Shapiro–Wilk test. Non-normal data underwent logarithmic transformation. A Fisher’s exact test was used to compare the occurrence of vomiting in cats with fPLI >5.3 and ⩽5.3 µg/l and in cats with TT4 ⩾45 and <45 nmol/l. To render testing more conservative and reduce the likelihood of false-positive results, the Bonferroni correction was used when multiple comparisons were made. Differences were considered significant at P <0.05.

Results

Number of samples, signalment, duration and mode of therapy, and differences between cat populations

Blood samples were available from 215 cats; 129 samples were from the Netherlands and 86 from Switzerland. Urine samples collected at home were available for 117 cats, 105 of which were from the Netherlands and 12 from Switzerland. Each cat was sampled only once. Age was known for 212/215 cats and ranged from 2–18 years (median 12 years), and body weight was obtained in 206/215 cats and ranged from 2.7–12.3 kg (median 5.5 kg). Sex was known for 213 cats; 139 (65%) were castrated males, and of the 74 female cats (35%), 71 were spayed. Breed was known in 202 cats; 165 (82%) were domestic shorthair cats and the remaining 37 were purebreds, including Norwegian Forest Cats (n = 9; 4.5%), Maine Coon (n = 8; 4%), Persian (n = 8; 4%), British Shorthair (n = 3; 1.5%), Siamese (n = 3; 1.5%), and one each (0.5%) of Burmese, Russian Blue, Birman, Ragdoll, Siberian and Somali cats.

Duration of diabetes was known for 76 of the cats (all from Switzerland) and ranged from 1–108 months (median 8 months). Type of insulin was known for 201 cats and consisted of porcine insulin zinc suspension in 161 cats (80%), insulin glargine in 39 cats (19.5%) and a human neutral protamine hagedorn insulin in one cat (0.5%). Insulin dosage ranged from 0.02–5.20 U/kg twice daily (median 0.5 U/kg). None of the cats was treated with non-insulin anti-diabetic drugs. Six cats were treated with diabetes diet alone. Type of food was noted in 189/215 cats; 101 cats (53%) received a diet designed specifically for diabetic cats, 53 cats (28%) were fed diets for the management of other diseases and 35 (18%) received regular commercial or home-made food. However, in most cases it was not clear whether a particular type of diet had been fed consistently.

The Swiss and Dutch cat populations did not differ with respect to sex (neutered males 60.3% and 69.8%, respectively; spayed females 39.7% and 30.2%, respectively [P = 0.057]) and breed (purebreds 11.5% and 22.1%, respectively; crossbreds 88.5% and 77.9%, respectively [P = 0.162]). Diabetic cats from Switzerland were younger than those from the Netherlands (median age 11 years [range 2–17 years] vs 13 years [range 4–18 years]; P = 0.004) and heavier (median body weight 5.9 kg [range 3.5–12.3 kg] vs 5.2 kg [range 2.7–11.6 kg]; P = 0.005). Regarding the type of insulin, diabetic cats from Switzerland received porcine insulin zinc suspension less often than those from the Netherlands (porcine insulin zinc suspension 55.8% and 95.9%, respectively; insulin glargine 44.2% and 4.1%, respectively [P <0.001]) and a lower dose of insulin (median insulin dose 0.46 IU/kg [range 0.02–1.56 IU/kg] vs 0.56 IU/kg [range 0.04–5.22 IU/kg]; P = 0.017).

IGF-1

IGF-1 concentration was measured in 202 cats; 122 were from the Netherlands and 80 from Switzerland. The concentrations ranged from 15–2470 ng/ml (median 583 ng/ml). The overall percentage of cats with IGF-1 concentrations >1000 ng/ml was 17.8% (36/202 cats). This percentage was 21.3% (26/122 cats) for the Netherlands and 12.5% (10/80 cats) for Switzerland. The prevalence of increased IGF-1 did not differ between the two countries (P = 0.110) (Figure 1).

Figure 1

Insulin-like growth factor 1 (IGF-1) concentrations (ng/ml) in 202 diabetic cats (All) treated at primary care veterinary clinics in the Netherlands (NL; n = 122) and in Switzerland (CH; n = 80). The bottom and top of a box represent the 25th and 75th percentiles, the line inside the box represents the median, the lines at the whiskers represent the 2.5th and 97.5th percentiles and the dots represent outliers

Two of the 36 cats with increased IGF-1 had clinical signs consistent with hypersomatotropism (large paw size in one cat and increased size of nasal bridge in the other); however, pituitary imaging was not feasible in these two cats. CT was performed in five cats with increased IGF-1, and a pituitary mass was identified in one. In one cat with IGF-1 concentrations slightly <1000 ng/ml, CT scanning revealed an enlarged pituitary.

Diabetic cats with IGF-1 concentrations >1000 ng/ml had significantly higher serum concentrations of creatinine than cats with IGF-1 ⩽1000 ng/ml (P = 0.036), but the remaining variables did not differ between cats with IGF-1 concentrations >1000 and ⩽1000 ng/ml.

Serum fPLI concentration

Serum fPLI concentration was measured in 196 cats; 118 were from the Netherlands and 78 from Switzerland. Concentrations ranged from 0.7–51.0 µg/l (median 4.4 µg/l). Eighty-six of the 196 cats (43.9%) had fPLI concentrations >5.3 µg/l. The percentage of cats with increased fPLI concentration was 45.8% (54/118 cats) for the Netherlands and 38.5% (30/78 cats) for Switzerland. The prevalence of increased fPLI concentration did not differ between the two countries (P = 0.312) (Figure 2).

Figure 2

Feline pancreas-specific lipase immunoreactivity (fPLI) concentrations (µg/l) in 196 diabetic cats (All) treated at primary care veterinary clinics in the Netherlands (NL; n = 118) and in Switzerland (CH; n = 78). Key to box plots as for Figure 1

Diabetic cats with fPLI concentrations >5.3 µg/l were significantly older than cats with fPLI concentrations ⩽5.3 µg/l (P = 0.002), had significantly higher serum concentrations of creatinine and activities of DGGR lipase and AST (P <0.001), and lower urine specific gravity (P = 0.022). The remaining variables did not differ between cats with fPLI concentrations >5.3 and ⩽5.3 µg/l.

TT4

TT4 was measured in 201 cats; 118 were from the Netherlands and 83 from Switzerland. The concentrations ranged from 6.4–143.0 nmol/l (median 23.2 nmol/l). Nine cats (4.5%) had TT4 concentrations ⩾45 nmol/l. The percentage of increased TT4 was 5.9% (7/118 cats) for the Netherlands and 2.4% (2/83 cats) for Switzerland. The prevalence of increased TT4 did not differ between the two countries (P = 0.234) (Figure 3). The diabetic cats with TT4 concentrations ⩾45 and <45 nmol/l did not differ with regard to any of the examined variables.

Figure 3

Total thyroxine (TT4) concentration (nmol/l) in 201 diabetic cats (All) treated at primary veterinary care clinics in the Netherlands (NL; n = 118) and in Switzerland (CH; n = 83). Key to box plots as for Figure 1

UCCR

UCCR was measured in 118 cats; 106 were from the Netherlands and 12 from Switzerland. The UCCR ranged from 0.4–139 x 10⁻⁶ (median, 17.6 × 10⁻⁶). Eighteen of the 118 cats (15.3%) had a UCCR >42 × 10⁻⁶. The percentage of increased UCCR was 15.1% (16/106 cats) for the Netherlands and 16.7% (2/12 cats) for Switzerland. The prevalence of increased UCCR did not differ between the two countries (P = 1.000) (Figure 4). Three of the 118 cats had clinical signs possibly associated with hypercortisolism (thin skin and some degree of alopecia).

Figure 4

Urinary corticoid-to-creatinine ratio (UCCR) in 118 diabetic cats (All) treated at primary veterinary care clinics in the Netherlands (NL; n = 106) and in Switzerland (CH; n = 12). Key to box plots as for Figure 1

Diabetic cats with a UCCR >42 × 10⁻⁶ had significantly higher urine protein-to-creatinine ratios than cats with UCCR ⩽42 × 10⁻⁶ (P <0.001). The other variables did not differ between cats with normal and increased UCCR.

Discussion

In the present study, IGF-1, fPLI, TT4 and UCCR were used as surrogate markers to estimate the prevalence of hypersomatotropism, pancreatitis, hyperthyroidism and hypercortisolism in cats with diabetes mellitus, respectively. For this purpose, cats were recruited from private veterinary clinics in the Netherlands and in Switzerland. The two cat populations differed with regard to age (lower in the cats from CH) and body weight (higher in the cats from CH). It could be speculated that the higher body weight in the cats from Switzerland contributed to an earlier onset of diabetes compared with the cats from the Netherlands. Obesity is a major risk factor for diabetes in cats, and overweight cats are several times more likely to develop diabetes than cats with a normal weight.¹⁶ The only other difference between the two cat populations was that the cats from the Netherlands were treated mainly with a porcine insulin preparation (in 96% of cases), whereas more than half of the cats from Switzerland (59%) received insulin glargine. We think that it is unlikely that these differences had a significant effect on the variables examined in this study. Of note, the prevalences of increased IGF-1, fPLI, TT4 and UCCR did not differ between the two cat populations.

Only cats that had been treated for at least 4 weeks were included in the study because unregulated glycaemic conditions could have led to false-negative or false-positive results. Serum IGF-1 concentrations may be low in untreated diabetic cats and increase into the reference interval within 4–8 weeks after initiating therapy.^10,17 Low serum IGF-1 concentrations also have been observed in newly diagnosed diabetic cats with hypersomatotropism before the start of insulin therapy.¹⁷ Similarly, TT4 may be low in untreated diabetic cats and increase during the first weeks of therapy. Thus, it is possible to miss a diagnosis of hyperthyroidism if TT4 is only evaluated at initial presentation.¹ With regard to the UCCR, it is known that the stress from non-adrenal disease may affect the hypothalamus–pituitary–adrenal (HPA) axis and lead to increased urinary cortisol excretion.^15,18 We hypothesised that the stress of the disease would decrease during treatment of diabetes leading to normalisation of the HPA axis. In cats with hyperthyroidism a decrease of UCCR during treatment has previously been shown.¹⁵

The percentage of cats with increased IGF-1 concentrations was lower in our study when compared with two previous studies from the UK (18% vs 32% and 26%).^5,19 It is conceivable that the prevalence of hypersomatotropism differs between countries. Although not statistically different, the prevalence of increased IGF-1 concentration was higher in the cats from the Netherlands (21%) than in the cats from Switzerland (13%). Also, comparison of studies from different countries can be affected by selection bias. We encouraged the participating veterinarians to collect samples from all diabetic cats presented for examination to avoid preferential sampling of cats that were difficult to regulate. Because we used increased IGF-1 concentration as a surrogate marker for hypersomatotropism, it is critical to determine to what degree the prevalence of increased IGF-1 concentrations reflect the true prevalence of hypersomatotropism. In recent years, measurement of IGF-1 has become a popular screening test for feline hypersomatotropism because it is considered a marker of growth hormone level in human patients, as well as in cats (www.thehormonelab.com).²⁰ However, similar to other tests, IGF-1 measurements may be associated with a certain rate of false-positive results. Circulating IGF-1 is almost completely bound to six different binding proteins, which are known to interfere with the assay. Although most immunoassays are preceded by methods that remove these binding proteins, the methods are not equally effective and constitute a potential source of error.^21,22

The assay used in our study was recently evaluated and results corresponded very well with those of an assay that uses acid chromatography and constitutes an accepted and reliable method for removal of binding proteins.¹² Therefore, we do not believe that technical problems are responsible for increased IGF-1 in our study. High IGF-1 concentrations in diabetic cats without hypersomatotropism have been reported by Lewitt et al.²³ A follow-up study showed that IGF-1 concentrations were significantly higher in diabetic cats that had received long-term insulin therapy than in cats on short- or medium-term insulin therapy and healthy cats.²⁴ Intensity of insulin therapy also seems to play a role. We have observed that diabetic cats receiving insulin via continuous rate infusion for 1 week had a more pronounced increase in IGF-1 concentrations after 6–8 weeks than cats receiving intermediate-acting insulin twice daily.^25,26 However, in the present study, duration of diabetes and insulin dose did not differ between cats with IGF-1 concentrations >1000 ng/ml and cats with IGF-1 concentrations ⩽1000 ng/ml, but this must be interpreted in view of the fact that the actual duration of disease was known in only approximately one-third of the cats.

According to the information submitted by the participating veterinarians, 34/36 cats with IGF-1 concentrations >1000 ng/ml did not have any clinical signs consistent with hypersomatotropism. However, physical changes caused by hypersomatotropism in cats have an insidious onset and progress slowly, and thus cats with early hypersomatotropism may be indistinguishable from other diabetic cats,^1,5,6,27 and a lack of clinical signs does not conclusively rule out the disease. Pituitary imaging is an essential tool for a definitive diagnosis of hypersomatotropism. In our study, CT imaging was carried out in 5/36 cats (14%) with IGF-1 concentrations >1000 ng/ml and only one of these cats had evidence of a pituitary tumour. It is conceivable that some of the remaining four cats had a tumour that was too small relative to the sensitivity of the CT scanner being used. We consider it unlikely, however, that this was the case in all four cats; ie, in some of the cats IGF-1 may have been false positive.

On the other hand, the chosen cut-off of 1000 ng/ml may have been too high, leading to underdiagnosis of hypersomatotropism. A case in point was one cat with a IGF-1 concentration <1000 ng/ml and a detectable pituitary mass, a scenario that also has previously been described by Niessen et al.⁶ There is no doubt that hypersomatotropism is an important disorder in cats. It is associated with variable degrees of insulin resistance, and diabetic cats that are difficult to regulate should be evaluated for hypersomatotropism. Further studies are needed to establish conclusively the prevalence of hypersomatotropism in different cat populations. In our study, cats with IGF-1 concentrations >1000 and those ⩽1000 ng/ml did not differ with regard to relevant clinical and laboratory variables (eg, body weight, blood glucose, serum fructosamine, insulin dose) and therefore we assume that not all cats with high IGF-1 concentrations had hypersomatotropism and that the prevalence of the disease was somewhat lower than 18%.

In humans, both acute and chronic pancreatitis are acknowledged causes of diabetes mellitus. Approximately 25% of humans develop diabetes mellitus within 5 years of an episode of acute pancreatitis. The mechanisms associated with this are speculative and include loss of β-cells, due to necrosis or immunological and metabolic factors.²⁸ The percentage is even higher in individuals with chronic pancreatitis; 40–70% develop diabetes mellitus within 10–20 years, mainly attributable to fibroinflammatory destruction of pancreatic β-cells.^29,30 On the other hand, humans with type 2 diabetes have a two- to three-fold risk of developing acute pancreatitis, possibly caused by factors such as obesity, hypertriglyceridaemia, proinflammatory cytokines and drugs.^31,32 Pancreatitis is considered an important disease in cats and is increasingly recognised. The potential association between pancreatitis and diabetes mellitus in cats is a matter of ongoing discussion, but because of the lack of long-term studies, it has not been clarified. In other words, it is not known whether pancreatitis in cats causes diabetes mellitus or vice versa.^8,33 Histological evidence of pancreatitis was found in a high percentage (51%) of diabetic cats.³⁴ However, a later study found that the percentage of clinically normal (non-diabetic) cats with histological evidence of pancreatitis (45%) was similar.^34,35 Further evidence for similar frequencies of pancreatitis in diabetic and non-diabetic cats was found in a recent study in which 68% of diabetic cats and 60% of matched control cats had histological evidence of pancreatitis.³⁶ In another study, the results of the serum fPLI measurements also indicated a high prevalence of pancreatitis with fPLI being increased in 33% of newly diagnosed diabetic cats and increased within 6 months of diagnosis in another 17% of diabetic cats. In the majority of cases, however, pancreatitis was considered to be subclinical because only 1/30 cats had a short episode of clinical signs possibly associated with pancreatitis.⁸

In the present study, 44% of cats had an increased fPLI concentration, which is in agreement with the results of a recent study at our clinic but in contrast to the findings of another study,^7,8 in which 83% of diabetic cats had increased fPLI concentrations. It is unclear whether these conflicting results reflect variances between countries or differences in study designs. Furthermore, Forcada et al⁷ found no association between age and fPLI, whereas in our study, cats with increased fPLI concentrations were significantly older. Body weight, clinical signs (including vomiting), blood glucose and fructosamine concentrations, as well as insulin dose, did not differ between cats with and without increased fPLI concentrations and, therefore, we assumed that pancreatic acinar damage was subclinical or very mild. The fPLI concentration and the DGGR lipase activity showed a highly significant correlation (r = 0.96, P = 0.000), confirming findings of recent studies.^37–39 It should be noted that the role of fPLI and DGGR lipase in diagnosing pancreatitis is a matter of controversy. Recently, it was shown that the agreement between both lipase assays and histological pancreatic inflammation is limited.³⁹

Hyperthyroidism is a very common endocrine disorder in cats and is one of the most frequently diagnosed disorders in elderly cats. There may be substantial differences in the prevalence of hyperthyroidism among geographical areas, and the prevalence of this disease has been shown to differ between referral and primary care hospitals.^40,41 In the latter, the prevalence of hyperthyroidism in elderly cats was 4% in Hong Kong, 9% in the UK and 11% in Berlin, Germany.^41–43 To the best of our knowledge, systematic studies on the prevalence of hyperthyroidism in cats with diabetes mellitus have not been performed and therefore it is not known whether the prevalence of this disease is higher in diabetic cats than in non-diabetic cats. Naturally occurring and experimentally induced hyperthyroidism result in impaired glucose tolerance, most likely attributable to an increase in peripheral insulin resistance.⁴⁴ It is therefore possible that hyperthyroidism contributes, to some extent, to the development of diabetes, although current data do not support this notion. In a retrospective case series, only 6% of diabetic cats were diagnosed with concurrent hyperthyroidism,⁹ which is similar to the prevalence of 5% found in our study. However, it should be noted that the diagnosis of hyperthyroidism may have been missed in some cats because of the presence of a concurrent disease lowering serum TT4 concentrations.⁴⁵

Studies on the use of UCCR for the diagnosis of hyperadrenocorticism in cats are scarce,^46,47 most likely because the disease is rare in cats. As in dogs, the stress associated with hospitalisation may cause a significant increase in UCCR in cats, and therefore, this parameter should be determined in urine samples collected by the owner at home.^15,48 In the present study, approximately half of the cat owners were able to provide ‘home’ samples. The percentage of increased UCCR was 15% and was clearly higher than expected. Clinical signs that may have been indicative of hypercortisolism were only present in 3/18 cats with an increased UCCR and were mild (ie, focal or multifocal alopecia, excessive dandruff). Additionally, none of the diabetes-related variables (eg, insulin dose, blood glucose and fructosamine concentrations) differed between cats with an increased UCCR and those with a normal ratio. Therefore, we assumed that the majority of cats with an increased UCCR did not have ‘true’ hypercortisolism, and that it was diabetes that caused increased glucocorticoid levels. Hyperactivity of the HPA axis has been shown in people with diabetes mellitus. People with diabetes mellitus may have significantly elevated levels of basal plasma cortisol, adrenocorticotropic hormone, total daily cortisol and 24 h urinary free cortisol, as well as a blunted suppression after dexamethasone administration.^49–52 To date, UCCR in diabetic cats has not been studied, but it is known that various non-adrenal diseases, including hyperthyroidism, affect the HPA axis and can lead to increases in UCCR.^15,18 In the study by de Lange et al,¹⁵ treatment of hyperthyroidism led to a marked decrease in UCCR and became normal in 6/7 cats. It would have been interesting to determine whether a similar decrease occurred in the present study after further treatment and improvement of glycaemic control. Unfortunately, follow-up examinations were not possible. Diabetic cats with elevated UCCR had significantly higher urine protein-to-creatinine ratios than cats with normal UCCR. It is known that hypercortisolism in dogs is associated with proteinuria.^53,54 This proteinuria may partly be due to cortisol-induced arterial hypertension. In line with this supposition, long-term hydrocortisone treatment has been reported to increase arterial blood pressure in dogs.⁵⁵

There were two main limitations of this study. The first one was a potential selection bias; cats with poorly regulated diabetes are more likely to be presented for veterinary care than well-regulated diabetic cats and therefore, concurrent diseases may be more prevalent than in the overall diabetic cat population. The second limitation was the lack of further investigation in most cases.

Conclusions

This is the first study to evaluate concentrations of IGF-1, fPLI, TT4 and UCCR simultaneously in cats with diabetes mellitus. Hyperthyroidism seems to be rare in diabetic cats, whereas an increased fPLI concentration, possibly reflecting pancreatitis, is common. As postulated in previous studies, pancreatitis appears to be subclinical in most cases. The high UCCR seen in a substantial percentage of cats may reflect activation of the HPA axis, which has been reported in human patients with diabetes. The percentage of cats with increased IGF-1 was rather high but lower than in recent studies. Longitudinal studies are needed to clarify whether it is possible to differentiate between cases in which pancreatitis is the cause or the consequence of diabetes, and what percentage of cats with increased IGF-1 but unremarkable pituitary imaging have, or will develop, hypersomatotropism.

Footnotes

Acknowledgements

We are grateful to Dr Barbara Contiero (Department of Animal Medicine, Production and Health, University of Padova, Italy) for statistical advice.

Accepted: 21 July 2016

Supplementary material

The following files are available:

Appendix 1: Questionnaire for owners of diabetic cats

Appendix 2: Physical examination undertaken by primary care veterinarians

Conflict of interest

JSS and JMS serve as director and associate director for research of the Gastrointestinal Laboratory (GI Lab) at Texas A&M University; JSS also serves as a paid consultant for IDEXX Laboratories. Both the GI Lab and IDEXX Laboratories perform fPLI testing on a fee-for-service basis. CER serves as consultant for Boehringer Ingelheim Vetmedica GmbH.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Reusch

. Feline diabetes mellitus. In: Feldman

Nelson

Reusch

, et al (eds). Canine and feline endocrinology. St Louis, MO: Elsevier, 2015, pp 258–314.

American Diabetes Association (ADA). Standards of medical care in diabetes – 2015. Diabetes Care 2015; S38: 1–93.

Rand

Fleeman

Farrow

, et al. Canine and feline diabetes mellitus: nature or nurture? J Nutr 2004; 134 Suppl 8: 2072S–2080S.

Scott-Moncrieff

. Insulin resistance in cats. Vet Clin North Am Small Anim Pract 2010; 40: 241–257.

Niessen

Petrie

Gaudiano

, et al. Feline acromegaly: an underdiagnosed endocrinopathy? J Vet Intern Med 2007; 21: 899–905.

Niessen

SJM

Forcada

Mantis

, et al. Studying cat (Felis catus) diabetes: beware of the acromegalic imposter. PLoS One 2015; 10: e0127794.

Forcada

German

Noble

, et al. Determination of serum fPLI concentrations in cats with diabetes mellitus. J Feline Med Surg 2008; 10: 480–487.

Zini

Hafner

Kook

, et al. Longitudinal evaluation of serum pancreatic enzymes and ultrasonographic findings in diabetic cats without clinically relevant pancreatitis at diagnosis. J Vet Intern Med 2015; 29: 589–596.

Crenshaw

Peterson

. Pretreatment clinical and laboratory evaluation of cats with diabetes mellitus: 104 cases (1992–1994). J Am Vet Med Assoc 1996; 209: 943–949.

10.

Alt

Kley

Tschuor

, et al. Evaluation of IGF-1 levels in cats with transient and permanent diabetes mellitus. Res Vet Sci 2007; 83: 331–335.

11.

Boretti

Sieber-Ruckstuhl

Gerber

, et al. Thyroid enlargement and its relationship to clinicopathological parameters and T4 status in suspected hyperthyroid cats. J Feline Med Surg 2009; 11: 286–292.

12.

Tschuor

Zini

Schellenberg

, et al. Evaluation of four methods used to measure plasma insulin-like growth factor 1 concentrations in healthy cats and cats with diabetes mellitus or other diseases. Am J Vet Res 2012; 73: 1925–1931.

13.

Forman

Shiroma

Armstrong

, et al. Evaluation of feline pancreas-specific lipase (SPEC fPL) for the diagnosis of feline pancreatitis. J Vet Intern Med 2009; 23: 733–734.

14.

Rijnberk

van Wees

Mol

. Assessment of two tests for the diagnosis of canine hyperadrenocorticism. Vet Rec 1988; 122: 178–180.

15.

de Lange

Galac

Trip

, et al. High urinary corticoid/creatinine ratios in cats with hyperthyroidism. J Vet Intern Med 2004; 18: 152–155.

16.

Scarlett

Donoghue

. Associations between body condition and disease in cats. J Am Vet Med Assoc 1998; 212: 1725–1731.

17.

Reusch

Kley

Casella

, et al. Measurements of growth hormone and insulin-like growth factor 1 in cats with diabetes mellitus. Vet Rec 2006; 158: 195–200.

18.

Henry

Clark

Young

, et al. Urine cortisol:creatinine ratio in healthy and sick cats. J Vet Intern Med 1996; 10: 123–126.

19.

Niessen

Church

Forcada

. Hypersomatotropism, acromegaly, and hyperadrenocorticism and feline diabetes mellitus. Vet Clin North Am Small Anim Pract 2013; 43: 319–350.

20.

Junnila

Strasburger

Bidlingmaier

. Pitfalls of insulin-like growth factor-I and growth hormone assays. Endocrinol Metab Clin North Am 2015; 44: 27–34.

21.

Chestnut

Quarmby

. Evaluation of total IGF-I assay methods using samples from type I and type II diabetic patients. J Immunol Methods 2002; 259: 11–24.

22.

Frystyk

Freda

Clemmons

. The current status of IGF-I assays – a 2009 update. Growth Horm IGF Res 2010; 20: 8–18.

23.

Lewitt

Hazel

Church

, et al. Regulation of insulin-like growth factor-binding protein-3 ternary complex in feline diabetes mellitus. J Endocrinol 2000; 166: 21–27.

24.

Starkey

Tan

Church

. Investigation of serum IGF-I levels amongst diabetic and non-diabetic cats. J Feline Med Surg 2004; 6: 149–155.

25.

Hafner

Dietiker-Moretti

Kaufmann

, et al. Intensive intravenous infusion of insulin in diabetic cats. J Vet Intern Med 2014; 28: 1753–1759.

26.

Reusch

Niessen

. Acromegaly: important disease or a distraction? Proceedings of the BSAVA Congress; 2014 April 3–6; Birmingham, UK.

27.

Niessen

. Feline acromegaly: an essential differential diagnosis for the difficult diabetic. J Feline Med Surg 2010; 12: 15–23.

28.

Das

SLM

Singh

Philips

ARJ

, et al. Newly diagnosed diabetes mellitus after acute pancreatitis: a systematic review and meta-analysis. Gut 2014; 63: 818–831.

29.

Czakó

Hegyi

Rakonczay

, et al. Interactions between the endocrine and exocrine pancreas and their clinical relevance. Pancreatology 2009; 9: 351–359.

30.

Rickels

Bellin

Toledo

FGS

, et al. Detection, evaluation and treatment of diabetes mellitus in chronic pancreatitis: recommendations from PancreasFest 2012. Pancreatology 2013; 13: 336–342.

31.

Girman

Kou

Cai

, et al. Patients with type 2 diabetes mellitus have higher risk for acute pancreatitis compared with those without diabetes. Diabetes Obes Metab 2010; 12: 766–771.

32.

Solanski

Barreto

Saccone

. Acute pancreatitis due to diabetes: the role of hyperglycaemia and insulin resistance. Pancreatology 2012; 12: 234–239.

33.

Davison

. Diabetes mellitus and pancreatitis – cause or effect? J Small Anim Pract 2015; 56: 50–59.

34.

Goossens

MMC

Nelson

Feldman

, et al. Response to insulin treatment and survival in 104 cats with diabetes mellitus (1985–1995). J Vet Intern Med 1998; 12: 1–6.

35.

De Cock

Forman

Farver

, et al. Prevalence and histopathologic characteristics of pancreatitis in cats. Vet Pathol 2007; 44: 39–49.

36.

Zini

Ferro

Lunardi

, et al. Exocrine pancreas in cats with diabetes mellitus. Vet Pathol 2016; 53: 145–152.

37.

Oppliger

Hartnack

Riond

, et al. Agreement of the serum Spec fPL ^TM and 1,2-o-dilauryl-rac-glycero-3-glutaric acid-(6’-methylresorufin) ester lipase assay for the determination of serum lipase in cats with suspicion of pancreatitis. J Vet Intern Med 2013; 27: 1077–1082.

38.

Oppliger

Hartnack

Reusch

, et al. Agreement of serum feline pancreas-specific lipase and colorimetric lipase assays with pancreatic ultrasonographic findings in cats with suspicion of pancreatitis: 161 cases (2008–2012). J Am Vet Med Assoc 2014; 244:1060–1065.

39.

Oppliger

Hilbe

Hartnack

, et al. Comparison of serum Spec fPL ™ and 1,2-o-dilauryl-rac-glycero-3-glutaric acid-(6’-methylresorufin) ester assay in 60 cats using standardized assessment of pancreatic histology. J Vet Intern Med 2016; 30: 764–770.

40.

McLean

Lobetti

Schoeman

. Worldwide prevalence and risk factors for feline hyperthyroidism: a review. J S Afr Vet Assoc 2014; 85: 1–6.

41.

Stephens

O’Neill

Church

, et al. Feline hyperthyroidism reported in primary-care veterinary practices in England: prevalence, associated factors and spatial distribution. Vet Rec 2014; 175: 458.

42.

De Wet

Mooney

Thompson

, et al. Prevalence of and risk factors for feline hyperthyroidism in Hong Kong. J Feline Med Surg 2009; 11: 315–321.

43.

Sassnau

. Epidemiological investigation on the prevalence of feline hyperthyroidism in an urban population in Germany. Tierärztl Prax 2006; 34: 450–457.

44.

Hoenig

Ferguson

. Impairment of glucose tolerance in hyperthyroid cats. J Endocrinol 1989; 121: 249–251.

45.

Scott Moncrieff

JCR

. Feline hyperthyroidism. In: Feldman

Nelson

Reusch

, et al (eds). Canine and feline endocrinology. St Louis, MO: Elsevier, 2015, pp 136–195.

46.

Goossens

MMC

Meyer

Voorhout

, et al. Urinary excretion of glucocorticoids in the diagnosis of hyperadrenocorticism in cats. Domest Anim Endocrinol 1995; 12: 355–362.

47.

Meij

Voorhout

Van Den Ingh

, et al. Trans-sphenoidal hypophysectomy for treatment of pituitary-dependent hyperadrenocorticism in 7 cats. Vet Surg 2001; 30: 72–86.

48.

Zimmer

Reusch

. Untersuchungen zum Kortisol-Kreatinin-Verhältnis im Urin (UCC) bei gesunden Katzen. Schweiz Arch Tierheilk 2003; 145: 323–328.

49.

Cameron

Kronfol

Greden

, et al. Hypothalamic-pituitary-adrenocortical activity in patients with diabetes mellitus. Arch Gen Psychiatry 1984; 41: 1090–1095.

50.

Carvalho

Urbanova

Hamer

, et al. Blunted glucocorticoid and mineralocorticoid sensitivity to stress in people with diabetes. Psychoneuroendocrinology 2015; 51: 209–218.

51.

Roy

Collier

Roy

. Hypothalamic-pituitary-adrenal axis dysregulation among diabetic outpatients. Psychiatry Res 1990; 31: 31–37.

52.

Roy

Gallucci

, et al. The ovine corticotropin-releasing hormone-stimulation test in type I diabetic patients and controls: suggestion of mild chronic hypercortisolism. Metabolism 1993; 42: 696–700.

53.

Ortega

Feldman

Nelson

, et al. Systemic arterial blood pressure and urine protein/creatinine ratio in dogs with hyperadrenocorticism. J Am Vet Med Assoc 1996; 209: 1724–1729.

54.

Smets

Lefebvre

Kooistra

, et al. Hypercortisolism affects glomerular and tubular function in dogs. Vet J 2012; 192: 532–534.

55.

Schellenberg

Mettler

Gentilini

, et al. The effects of hydrocortisone on systemic arterial blood pressure and urinary protein excretion in dogs. J Vet Intern Med 2008; 22: 273–281.