The potential role of leptin and adiponectin in obesity: a comparative review

Leptin and adiponectin are adipokines produced by the white adipose tissue. The adipokines have been shown to be valuable quantitative markers of adiposity in dogs. Leptin positively correlates with body condition score (BCS) in dogs, regardless of age, sex and breed, and is influenced by feeding state, pharmacological treatment and thyroid gland activity. Conversely, adiponectin negatively correlates with body fat mass and is therefore more abundant in lean animals. The implication of leptin and adiponectin in the pathogenesis of metabolic syndrome is well established in humans, but currently lacking in dogs. Additional studies are necessary to demonstrate their potential usefulness for monitoring the progression of obesity-related diseases and response to treatment. To date, measurement of canine leptin and adiponectin has been used in experimental studies only, whereas bodyweight and BCS are considered the first-approach parameters for the routine assessment of body fat content in obese dogs.     

Preliminary report on the expression of leptin and leptin receptor (ObR) in normal, hyperplastic and neoplastic canine mammary tissues

Leptin and its receptor (ObR) expression were investigated by immunohistochemistry in normal, hyperplastic and neoplastic canine mammary tissues and related to clinical-pathological features. Leptin expression was detected in healthy mammary tissues, adenosis and in benign mammary tumours and was lower in ductal hyperplasias and malignant tumours. A high percentage of ObR-positive cells were present in adenosis, benign tumours and in complex carcinomas, while ObR expression was lower in healthy mammary tissues, in ductal hyperplasias and in a large part of invasive mammary carcinomas. Our data demonstrated that cancer cells expressed at low level leptin and ObR in canine mammary tumours with a more aggressive behaviour, as well as in lymph node metastases. Consequently, leptin and ObR expressions in this species resulted to be not associated with a reduced overall survival.     

Compensation for obesity-induced insulin resistance in dogs: assessment of the effects of leptin, adiponectin, and glucagon-like peptide-1 using path analysis

The hormonal mediators of obesity-induced insulin resistance and compensatory hyperinsulinemia in dogs have not been identified. Plasma samples were obtained after a 24-h fast from 104 client-owned lean, overweight, and obese dogs. Plasma glucose and insulin concentrations were used to calculate insulin sensitivity and β-cell function with the use of the homeostasis model assessment (HOMA_{insulin sensitivity} and HOMA_{β-cell function}, respectively). Path analysis with multivariable linear regression was used to identify whether fasting plasma leptin, adiponectin, or glucagon-like peptide-1 concentrations were associated with adiposity, insulin sensitivity, and basal insulin secretion. None of the dogs were hyperglycemic. In the final path model, adiposity was positively associated with leptin (P < 0.01) and glucagon-like peptide-1 (P = 0.04) concentrations. No significant total effect of adiposity on adiponectin in dogs (P = 0.24) was observed. If there is a direct effect of leptin on adiponectin, then our results indicate that this is a positive relationship, which at least partly counters a negative direct relationship between adiposity and adiponectin. Fasting plasma leptin concentration was directly negatively associated with fasting insulin sensitivity (P = 0.01) and positively associated with β-cell function (P < 0.01), but no direct association was observed between adiponectin concentration and either insulin sensitivity or β-cell function (P = 0.42 and 0.11, respectively). We conclude that dogs compensate effectively for obesity-induced insulin resistance. Fasting plasma leptin concentrations appear to be associated with obesity-associated changes in insulin sensitivity and compensatory hyperinsulinemia in naturally occurring obese dogs. Adiponectin does not appear to be involved in the pathophysiology of obesity-associated changes in insulin sensitivity.     

Serum antibody responses to vaccinal antigens in lean and obese geriatric dogs

The immune responses in control dogs [1 to 4 years of age, body condition score (BCS): 4 to 5 out of 9] were compared to those of aging dogs (based on breed and body size) either categorized as lean (BCS: 4 to 5 out of 9) or obese (BCS: 8 to 9 out of 9). Of interest were the serum titers to the following common agents found in vaccines, canine parainfluenza virus (CPIV), canine parvovirus (CPV), canine distemper virus (CDV), canine respiratory coronavirus (CRCoV), and Bordetella bronchiseptica. There were no statistical differences in the antibodies to CPIV, B. bronchispetica, and CRCoV, among the age/weight categories, nor among the age/weight categories and the time, in days, between the date of sample collection and the date of the last recorded vaccination for CPIV, B. bronchiseptica, CPV, and CDV. For CPV, the control dogs had significantly (P < 0.002) higher serum neutralization (SN) titers than the lean geriatric dogs and the obese geriatric dogs. For CDV SN titers, the only statistically significant (P = 0.01) difference was that the control dogs had higher SN titers than the lean geriatric dogs.     

Distinct adiponectin profiles might contribute to differences in susceptibility to type 2 diabetes in dogs and humans

Dogs develop obesity-associated insulin resistance but not type 2 diabetes mellitus. Low adiponectin is associated with progression to type 2 diabetes in obese humans. The aims of this study were to compare total and high molecular weight (HMW) adiponectin and the ratio of HMW to total adiponectin (S_A) between dogs and humans and to examine whether total or HMW adiponectin or both are associated with insulin resistance in naturally occurring obese dogs. We compared adiponectin profiles between 10 lean dogs and 10 lean humans and between 6 lean dogs and 6 age- and sex-matched, client-owned obese dogs. Total adiponectin was measured with assays validated in each species. We measured S_A with velocity centrifugation on sucrose gradients. The effect of total and HMW adiponectin concentrations on MINMOD-estimated insulin sensitivity was assessed with linear regression. Lean dogs had total and HMW adiponectin concentrations three to four times higher than lean humans (total: dogs 32 ± 5.6 mg/L, humans 10 ± 1.3 mg/L, P<0.001; HMW: dogs 25 ± 4.5 mg/L, humans 6 ± 1.3 mg/L, P<0.001) and a higher S_A (dogs: 0.78 ± 0.05; humans: 0.54 ± 0.08, P = 0.002). Adiponectin concentrations and S_A were not lower in obese dogs (0.76 ± 0.05 in both groups; P=1). Total adiponectin, HMW adiponectin, and S_A were not associated with insulin sensitivity in dogs. We propose that differences in adiponectin profiles between humans and dogs might contribute to the propensity of humans but not dogs to develop type 2 diabetes. Dogs with chronic, naturally occurring obesity do not have selectively reduced HMW adiponectin, and adiponectin does not appear to be important in the development of canine obesity-associated insulin resistance.     

Canine adiponectin: cDNA structure, mRNA expression in adipose tissues and reduced plasma levels in obesity

Adiponectin is a protein synthesized and secreted by adipocytes. Decreased adiponectin is responsible for insulin resistance and atherosclerosis associated with human obesity. We obtained a cDNA clone corresponding to canine adiponectin, whose nucleotide and deduced amino acid sequences were highly identical to those of other species. Adiponectin mRNA was detected in adipose tissues, but not in other tissues, of dogs. When 22 adult beagles were given a high-energy diet for 14 weeks, they became obese, showing heavier body weights, higher plasma leptin concentrations, but lower plasma adiponectin concentrations. The adiponectin concentrations of plasma samples collected from 71 dogs visiting veterinary practices were negatively correlated to plasma leptin concentrations, being lower in obese than non-obese dogs. These results are compatible with those reported in other species, and suggest that adiponectin is an index of adiposity and a target molecule for studies on diseases associated with obesity in dogs.     

Validation of 2 commercially available enzyme-linked immunosorbent assays for adiponectin determination in canine serum samples

The aim of this study was to validate 2 commercially available enzyme-linked immunosorbent assays (ELISAs) for adiponectin in dogs, 1 canine-specific and 1 originally designed for measurements in humans. Intra-assay and interassay precision was evaluated by multiple measurements in canine serum samples, and assay accuracy was indirectly determined by linearity under dilution. Interference caused by hemolysis and lipemia was also studied. Both assays were subsequently used for measuring adiponectin concentrations in clinically healthy dogs and those with different grades of obesity. The intra-assay and inter-assay precision was less than 7.5% and 13.5% in serum samples with low and high adiponectin concentrations, respectively. Lipemia and hemolysis did not affect the results of any of the assays. Both assays were able to differentiate lean dogs from those that were overweight or obese on the basis of the measured adiponectin concentrations. From these results it can be concluded that canine adiponectin concentrations can be measured reliably by means of the 2 ELISAs evaluated in this study.     

Evaluation of high-molecular weight adiponectin in horses

Objective-To characterize adiponectin protein complexes in lean and obese horses. Animals-26 lean horses and 18 obese horses. Procedures-Body condition score (BCS) and serum insulin activity were measured for each horse. Denaturing and native western blot analyses were used to evaluate adiponectin complexes in serum. A human ELISA kit was validated and used to quantify high-molecular weight (HMW) complexes. Correlations between variables were made, and HMW values were compared between groups. Results-Adiponectin was present as a multimer consisting of HMW (> 720-kDa), low-molecular weight (180-kDa), and trimeric (90-kDa) complexes in serum. All complexes were qualitatively reduced in obese horses versus lean horses, but the percentage of complexes < 250 kDa was higher in obese versus lean horses. High-molecular weight adiponectin concentration measured via ELISA was negatively correlated with serum insulin activity and BCS and was lower in obese horses (mean ± SD, 3.6 ± 3.9 μg/mL), compared with lean horses (8.0 ± 4.6 μg/mL). Conclusions and Clinical Relevance-HMW adiponectin is measurable via ELISA, and concentration is negatively correlated with BCS and serum insulin activity in horses. A greater understanding of the role of adiponectin in equine metabolism will provide insight into the pathophysiology of metabolic disease conditions.     

Adiposity and adiponectin in dogs: investigation of causes of discrepant results between two studies

Although one study showed lower adiponectin concentrations in obese dogs, other recent studies indicate that adiponectin might not be decreased in obese dogs, raising the possibility that the physiology of adiponectin is different in dogs than in humans. The aim of this study was to investigate possible causes of the discrepancy between the two largest studies to date that assessed the association between adiposity and adiponectin concentration in dogs, including the validity of the assay, laboratory error, and the effects of breed, sex, and neuter status on the relationship between adiposity and adiponectin concentrations. Adiponectin concentrations measured with a previously validated adiponectin ELISA were compared with those estimated by Western blotting analysis of reduced and denatured plasma samples. The possibility of laboratory error and the effect of EDTA anticoagulant and aprotinin were tested. Adiponectin concentration was measured by ELISA in 20 lean dogs (10 male and 10 female, 5 neutered in each sex). There was close correlation between adiponectin concentrations measured by ELISA and those estimated by Western blotting analysis (r = 0.90; P < 0.001). There was no substantial effect of EDTA, aprotinin, or laboratory error on the results. There was confounding by neuter status of the relationship between adiposity and adiponectin concentrations, but adiponectin concentrations were not significantly lower in male than in female lean dogs (females, 36 mg/L; males, 26 mg/L; P > 0.20) and were not significantly lower in intact than in neutered lean male dogs (intact, 28 mg/L; neutered, 23 mg/L; P = 0.49). We conclude that the adiponectin ELISA previously validated for use in dogs appears to be suitable for determination of canine adiponectin concentrations and that testosterone does not appear to have a strong effect on plasma adiponectin concentrations in dogs. Obesity might decrease adiponectin concentrations in intact but not in neutered dogs.     

mRNA expression of canine ATP10C, a P4-type ATPase, is positively associated with body condition score

Mouse and human Atp10c genes are strong candidates for changes in bodyweight and glucose homeostasis. Using comparative genomic analysis, a novel canine P4-type ATPase, ATP10C, was identified. Expression of ATP10C was compared between sex-matched lean (body condition score, BCS<8; n=7) and obese (BCS?8, n=8) client-owned dogs of comparable ages. Canine ATP10C is highly expressed in visceral and subcutaneous fat at approximately 3-fold levels compared to the omental adipose depot. There was a 5-fold significant increase (P<0.0001) in mRNA expression of ATP10C in dogs with a BCS?8.     

Experimental and clinical studies on plasma leptin in obese dogs

Leptin is a protein synthesized and secreted primarily by adipocytes, and the circulating leptin concentration is elevated in obese humans and rodents. Recently, we have established a sandwich enzyme-linked immunosorbent assay for canine leptin. In the present study, plasma leptin concentrations were measured in experimentally developed obese beagles and in clinically obese dogs. When 5 male beagles were given a high-energy diet for 3 months, all of them became obese and the plasma leptin concentration significantly increased from 2.4+/-1.2 to 4.9+/-0.9 ng/ml, positively correlating with body fat content estimated by the deuterium oxide dilution method (r=0.87). The leptin concentrations of plasma samples collected from 59 dogs in veterinary practices were compared with their body condition scores (BCS). The plasma leptin concentrations of obese dogs were 9.7+/-0.7 and 12.3+/-1.5 ng/ml at BCS=4 and BCS=5, respectively, which were significantly higher than those of optimal (BCS=3) dogs (2.7+/-0.3 ng/ml). There was no significant effect of sex and breed. A weak positive correlation (r=0.37) was found between the plasma leptin concentration and age, probably due to the lesser content of visceral fat in puppies younger than 1 year old. These results indicate that plasma leptin is a good index of adiposity in dogs regardless of breed, age and sex, and may be useful for quantitative assessment of obesity in small animal practice.     

Prevalence of obese dogs in a population of dogs with cancer

OBJECTIVE: To determine the body condition score (BCS) distribution for dogs examined at a teaching hospital and examine whether the BCS distribution for dogs with cancer differed significantly from the distribution for dogs without cancer. SAMPLE POPULATION: 1,777 dogs with cancer and 12,893 dogs without cancer. PROCEDURES: A retrospective prevalence case-control study was conducted that used medical records from 1999 to 2004. Information was collected on BCS (9-point system), age, breed, sex, neuter status, diagnosis, and corticosteroid administration. Body condition score at the time of examination for cancer (dogs with cancer) or first chronologic visit (dogs without cancer) was recorded. Logistic regression was used to compare BCS prevalence distributions between groups. RESULTS: The overall prevalence of obese dogs (BCS >or= 7/9) was 14.8% (2,169/14,670), and the overall prevalence of overweight dogs (BCS >or= 6/9 to < 7/9) was 21.6% (3,174/14,670). There was a significant difference in the BCS distribution between dogs with and without cancer, with a slightly lower prevalence of being overweight and obese in dogs with cancer. The prevalence of obese and overweight dogs varied with specific cancer types when compared with the prevalence for dogs without cancer. CONCLUSIONS AND CLINICAL RELEVANCE: Differences in obesity prevalence among cancer types is suggestive of an incongruous effect of this variable on cancer expression or a differential effect of specific cancer types on weight status. Systematic use of BCSs will help elucidate the association between obesity and cancer development.     

Correlation of Foxp3 positive regulatory T cells with prognostic factors in canine mammary carcinomas

Regulatory T cells (Treg) cells play a crucial role in tumor progression by suppressing anti-tumor immunity, but are not well-documented in veterinary oncology. To identify the characteristics of Treg cells in tumor microenvironments, the numbers of Treg cells were analyzed and compared with histological prognostic factors and molecular biomarkers in canine mammary carcinoma (MC) tissues (n=37). Abundant Treg cells were associated with high histological grade and lymphatic invasion. The numbers of Treg cells infiltrating intratumoral areas markedly increased in tumors with poor prognostic factors, such as high histological grade, lymphatic invasion, and necrosis. These findings suggest that Treg cells play a role in canine MC progression. Furthermore, Treg cell numbers in intratumoral compartments may provide a potential prognostic factor when assessing canine MCs, which may in turn lead to the development of new immunologic therapeutics.     

Obesity induced changes to plasma adiponectin concentration and cholesterol lipoprotein composition profile in cats

Feline obesity generally results in aberrations to plasma metabolite levels, such as lipid concentrations and lipoprotein composition. This study sought to investigate the resultant effect of obesity on cholesterol lipoprotein composition and circulating adiponectin concentrations in cats. Plasma glucose, lipids (triglyceride, cholesterol and free fatty acid), insulin and adiponectin concentrations, and cholesterol lipoprotein composition were measured and compared between body condition score (BCS) determined normal healthy control and obese cats. Although the obese group demonstrated higher levels of plasma cholesterol, glucose, and triglycerides, as compared to healthy controls, the difference was insignificant thus indicating that the BCS determined obese cats may have been overweight and not morbidly obese. Plasma insulin levels were significantly higher (25–30%) versus healthy control animals thereby possibly hinting at the ensuing emergence of obesity induced insulin resistance. However, the BCS determined obese cat demonstrated a significant reduction (p < 0.05) in plasma adiponectin concentration and a significant increase (p < 0.05) in LDL-cholesterol % as compared to age matched healthy control animals. This would indicate that changes in plasma adiponectin concentration and cholesterol lipoprotein composition may be good early indicators of obesity in cats.     

The effect of body condition on propofol requirement in dogs

ObjectiveTo determine if body condition score (BCS) influences the sedative effect of intramuscular (IM) premedication or the dose of intravenous (IV) propofol required to achieve endotracheal intubation in dogs.Study designProspective clinical study.AnimalsForty–six client–owned dogs undergoing general anaesthesia.MethodsDogs were allocated to groups according to their BCS (BCS, 1 [emaciated] to 9 [obese]): Normal–weight Group (NG, n = 25) if BCS 4–5 or Over–weight Group (OG, n = 21) if BCS over 6. Dogs were scored for sedation prior to IM injection of medetomidine (5 μg kg^?1) and butorphanol (0.2 mg kg^?1) and twenty minutes later anaesthesia was induced by a slow infusion of propofol at 1.5 mg kg^?1 minute^?1 until endotracheal intubation could be achieved. The total dose of propofol administered was recorded. Data were tested for normality then analyzed using Student t–tests, Mann–Whitney U tests, chi–square tests or linear regression as appropriate.ResultsMean ( ± SD) propofol requirement in NG was 2.24 ± 0.53 mg kg^?1 and in OG was 1.83 ± 0.36 mg kg^?1. The difference between the groups was statistically significant (p = 0.005). The degree of sedation was not different between the groups (p = 0.7). Post–induction apnoea occurred in 11 of 25 animals in the NG and three of 21 in OG (p = 0.052).ConclusionsOverweight dogs required a lower IV propofol dose per kg of total body mass to allow tracheal intubation than did normal body condition score animals suggesting that IV anaesthetic doses should be calculated according to lean body mass. The lower dose per kg of total body mass may have resulted in less post–induction apnoea in overweight/obese dogs. The effect of IM premedication was not significantly affected by the BCS.Clinical relevanceInduction of general anaesthesia with propofol in overweight dogs may be expected at lower doses than normal–weight animals.     

Association of Obesity with Serum Leptin,Adiponectin, and Serotonin and Gut Microflora in Beagle Dogs

Background

Serotonin (5‐hydroxytryptamine, 5HT) is involved in hypothalamic regulation of energy consumption. Also, the gut microbiome can influence neuronal signaling to the brain through vagal afferent neurons. Therefore, serotonin concentrations in the central nervous system and the composition of the microbiota can be related to obesity.

Objective

To examine adipokine, and, serotonin concentrations, and the gut microbiota in lean dogs and dogs with experimentally induced obesity.

Animals

Fourteen healthy Beagle dogs were used in this study.

Methods

Seven Beagle dogs in the obese group were fed commercial food ad libitum, over a period of 6 months to increase their weight and seven Beagle dogs in lean group were fed a restricted amount of the same diet to maintain optimal body condition over a period of 6 months. Peripheral leptin, adiponectin, 5HT, and cerebrospinal fluid (CSF‐5HT) levels were measured by ELISA. Fecal samples were collected in lean and obese groups 6 months after obesity was induced. Targeted pyrosequencing of the 16S rRNA gene was performed using a Genome Sequencer FLX plus system.

Results

Leptin concentrations were higher in the obese group (1.98 ± 1.00) compared to those of the lean group (1.12 ± 0.07, P = .025). Adiponectin and 5‐hydroytryptamine of cerebrospinal fluid (CSF‐5HT) concentrations were higher in the lean group (27.1 ± 7.28) than in the obese group (14.4 ± 5.40, P = .018). Analysis of the microbiome revealed that the diversity of the microbial community was lower in the obese group. Microbes from the phylum Firmicutes (85%) were predominant group in the gut microbiota of lean dogs. However, bacteria from the phylum Proteobacteria (76%) were the predominant group in the gut microbiota of dogs in the obese group.

Conclusions and Clinical Importance

Decreased 5HT levels in obese group might increase the risk of obesity because of increased appetite. Microflora enriched with gram‐negative might be related with chronic inflammation status in obese dogs.     

Increased serum concentrations of adiponectin in canine hypothyroidism

Serum concentrations of adiponectin were compared between sex-matched hypothyroid (n = 18) and euthyroid (n = 18) client-owned dogs with comparable ages and body condition scores (BCS). Concentrations of adiponectin (mean; 95% confidence interval) were significantly (P < 0.01) higher in hypothyroid (17.2 µg/mL; 12.1–20.5 µg/mL) than healthy (8.0 µg/mL; 5.6–11.4 µg/mL) dogs following adjustment for potential confounders (BCS, age and sex). Serum concentrations of adiponectin were significantly negatively associated with concentrations of total thyroxine (P <0.05) and positively correlated with concentrations of cholesterol (r = 0.6, P <0.01) in hypothyroid dogs. In conclusion, this study demonstrated increased serum concentrations of adiponectin in dogs with hypothyroidism. Suggestive of the presence of resistance to adiponectin that could have contributed to development of hyperlipidemia and insulin resistance in these dogs or alternatively, could be a consequence of these metabolic alterations.     

Plasma biomarkers profile of female dogs with mammary carcinoma and its association with clinical and pathological features

The immunological biomarkers profiles were evaluated using Luminex as putative measures to monitor canine mammary carcinomas (MCs). Forty female dogs were categorized into benign mixed tumour (MC‐BMT = 28) and mammary carcinoma (MC=12). The ascendant biomarker signatures were used to compare the groups. For example, a higher frequency of MC‐BMT animals producing IL‐6, CXCL‐8 and CXCL‐10 was observed, whereas for the MC group IL‐2 and CXCL‐8 were detected. MC‐BMT animals without metastasis had an increase in the levels of IL‐2, CXCL‐8, CXCL‐10, IL‐6, TNF‐α, IL‐15 and a decrease in IL‐10 and CXCL‐8. MC‐BMT animals with metastasis showed only an increase in CXCL‐10 and a decrease in IL‐18. After comparing the ascendant signatures following the presence of metastasis in both groups, a higher frequency of dogs exhibiting IL‐10 production was observed. Pearson correlation (P = 0.0273) and receiver operating characteristic (ROC) curve analysis revealed that this pattern was associated with worse outcome and lower survival rates in MC animals.     

Association of canine obesity with reduced serum levels of C-reactive protein.

The prevalence of obesity is increasing in dogs as well as in humans. C-reactive protein (CRP) is an important tool for the detection of inflammation and/or early tissue damage and is linked to obesity in humans. The objective of the present study was to determine if serum CRP levels are altered in obese dogs. Fifteen lean (control group) and 16 overweight (obese group) dogs were examined. Blood samples were collected under fasted conditions for serum determination of CRP, glucose, insulin, cholesterol, triglyceride, and fructosamine. Results indicated that obese dogs were insulin resistant because serum insulin and insulin/glucose ratios were higher than in lean dogs (P < or = 0.05). Serum CRP concentrations were lower in obese dogs than in controls (P < or = 0.001). C-reactive protein was negatively correlated with insulin/glucose ratio (R = -0.42) and cholesterol (R = -0.39; P < or = 0.05). Furthermore, levels of cholesterol, triglycerides, and fructosamine were increased in the obese group compared with the control group. Based on these results, it can be postulated that CRP production is inhibited by obesity and insulin resistance in dogs.     

Obesity induced changes to plasma adiponectin concentration and cholesterol lipoprotein composition profile in cats

Feline obesity generally results in aberrations to plasma metabolite levels, such as lipid concentrations and lipoprotein composition. This study sought to investigate the resultant effect of obesity on cholesterol lipoprotein composition and circulating adiponectin concentrations in cats. Plasma glucose, lipids (triglyceride, cholesterol and free fatty acid), insulin and adiponectin concentrations, and cholesterol lipoprotein composition were measured and compared between body condition score (BCS) determined normal healthy control and obese cats. Although the obese group demonstrated higher levels of plasma cholesterol, glucose, and triglycerides, as compared to healthy controls, the difference was insignificant thus indicating that the BCS determined obese cats may have been overweight and not morbidly obese. Plasma insulin levels were significantly higher (25–30%) versus healthy control animals thereby possibly hinting at the ensuing emergence of obesity induced insulin resistance. However, the BCS determined obese cat demonstrated a significant reduction (p < 0.05) in plasma adiponectin concentration and a significant increase (p < 0.05) in LDL-cholesterol % as compared to age matched healthy control animals. This would indicate that changes in plasma adiponectin concentration and cholesterol lipoprotein composition may be good early indicators of obesity in cats.