Materials and Methods

Animals—Twenty European domestic shorthair adult neutered male and female colony cats with a mean ± SD age of 4.6 ± 0.9 years were used for this study. All cats were classified as overweight on the basis of a BCS ranging from 7 to 9 on a 9-point scale (1 to 3 = lean; 4 to 5 = ideal; 6 to 9 = overweight to obese).³² Overweight cats had been overweight for > 1 year prior to the beginning of the study. Weight gain had been originally induced by allowing ad libitum food consumption. All cats were judged healthy on the basis of results of physical examination and clinical laboratory data.

Cats were maintained at the Affinity Petcare facility in standard colony conditions. They were housed by groups (n = 6 or 7/group) in indoor heated parks (6 × 12 m), with 12 hours of light daily (from 7 am to 7 pm); free access to water was provided. All animal handling procedures were carried out following the recommendations of the Spanish Law on Animal Welfare No. 22/2003.

Diet—Cats were allotted to 1 of 3 diet groups (n = 6 or 7/group) to balance age, sex, and BCS distributions among groups. During a run-in period of 2 months, all cats were adapted to a maintenance diet. First samples were collected after this adaptation period (T₀). After first sample collection, the 3 groups of cats were randomly allocated to 1 of 3 moderate-fat, moderate-energy premium diets: SF, a regimen with 3.4% beef tallow as a source of saturated fat; SF+C, a regimen with 3.4% beef tallow and citrus flavanones (0.03% hesperidin, 0.017% narangin)^b; and OO, a regimen with 3.4% olive oil as a source of MUFAs (Appendix).³³ Dietary composition was chemically analyzed, and metabolizable energy was measured in vivo (AAFCO protocol³⁴ and National Research Council equation³⁵). Each diet was formulated as kibble, in accordance with the FEDIAF (European Pet Food Industry Federation)³⁶ nutrient guide for cats, and balanced to meet adult maintenance requirements.

Cats were fed the test diets for 5 months. The composition of diets fed was unknown to investigators involved in conducting the experiment or acquiring data. During the whole study, food in excess of that usually consumed by the cats was provided every morning and was available for 24 h/d. The next morning, the remaining food was weighed, mean daily food intake by group was recorded, and fresh food was distributed. Mean daily food and energy intake are expressed in kilograms of BW per group of cats (calculated as total amount of food or energy consumed by each group, divided by the combined BW of the whole group). Body weight (kg) was recorded once a week in the morning before food distribution.

Experimental protocol—Blood samples were obtained via jugular venipuncture (0.6 × 25 mm, 23-gauge needle) at T₀ and 1 (T₁), 3 (T₂), and 5 (T₃) months after test diet initiation. Food was withheld from cats 20 hours before sample collection, and collected blood was immediately transferred into EDTA- and heparincoated tubes and stored on ice for serum or plasma harvesting. Urine samples were obtained at T₀ and T₃ via urinary catheter from sedated cats (100 μg of medetomidine^d/kg, IM).

Sample processing—After collection, blood samples were centrifuged (1,500 × g at 4°C for 10 minutes) and harvested serum or plasma samples were frozen before being sent on dry ice to the laboratory. At the laboratory, samples were stored at —20°C until assays were performed. A complete serum or plasma biochemical analysis was performed by a veterinary diagnostic laboratory/ Plasma cholesterol and TAG concentrations were measured via enzymatic colorimetry. Urine samples were centrifuged (1,500 × g at 4°C for 10 minutes) prior to storage at −80°C.

Measurement of urinary F2-isoprostane concentration—F2-isoprostane concentration was determined in all urine samples via 1 enzyme immunoassay^f at the end of the study.¹³ Briefly, urine samples were thawed, and urine pH was adjusted to 2 to 2.5 with 2N HCl. Samples were centrifuged for 10 minutes at 1,500 × g and 4°C. One milliliter of supernatant was run through a C 1-mL, 100-mg purification cartridge column.^g The column was rinsed with 1 mL of water and 1 mL of hexane. Elution was performed with 2 mL (4 × 500 μL) of ethyl acetate-methanol (95:5 [vol/vol]). Eluate was vacuum dried and reconstituted with buffer provided by the manufacturer. The kit was subsequently used in accordance with the manufacturer's instructions. The within-assay coefficient of variation was 7%; the between-assay coefficient of variation was 17%. Urinary creatinine concentration was determined by spectrophotometry in a veterinary laboratory.^e F2-isoprostane concentrations were standardized for urine dilution by expressing data as an F2-isoprostane-to-urinary creatinine concentration ratio.³⁷

Statistical analysis—Data were evaluated for normality of distribution^h and equality of variance.ⁱ To assess whether diet groups differed in values of serum or plasma and urinary variables at T₀, general linear models^j were used. Then, to assess the effects of diet, time, and their interaction on the variables, multivariate ANOVA^j was performed. Age and sex effects were also included in the models because these variables can influence results of serum or plasma biochemical analyses. When unequal variances were detected, data were logarithmically transformed; however, because the transformed and untransformed data resulted in qualitatively similar results, the mean ± SEM values for untransformed data are reported. For comparisons of food intake among diet groups, the mean daily food and energy intake per kilogram of BW was calculated for each group and compared.^j All comparisons were preplanned. For all analyses, a value of P < 0.05 was considered significant.

Results

Animals—Mean age, BW, and BCS of cats at the beginning of the study did not differ significantly among the 3 diet groups (SF, SF+C, and OO) of overweight cats (Table 1). Mean food and energy intake per group was significantly higher in the SF group (14.5 ± 0.07 g/kg of BW and 55 ± 0.3 kcal/kg of BW, respectively) than in the OO group (14.1 ± 0.09 g/kg of BW [P = 0.004] and 53 ± 0.3 kcal/kg of BW [P = 0.003], respectively) and the SF+C group (13.6 ± 0.15 g/kg of BW [P = 0.002] and 51 ± 0.6 kcal/kg of BW [P = 0.001], respectively).

Consumption of none of the moderate-energy diets resulted in significant change in BW from that before diet transitioning (T₀) after 5 months of ad libitum feeding (T₃; Table 2).

No difference was evident between the 3 groups of overweight cats at T₀ with respect to results of serum or plasma and urinary F2-isoprostane assays (Table 3). Serum or plasma biochemical values remained within reference limits during the entire study period. A significant (P = 0.02) effect of time on plasma cholesterol concentration was detected. The SF+C group was the only one to have a significant decrease from the T₀ value in cholesterol concentration at T₁ (1 month after diet initiation; P = 0.04), T₂ (3 months after diet initiation; P = 0.01), and T₃ (5 months after diet initiation; P = 0.01).

A significant effect of diet was also evident for plasma cholesterol concentration (P = 0.02; the SF+C group had a lower value than did the SF [P = 0.04] or OO [P = 0.003] groups) and TAG concentration (P = 0.03; the SF+C group had a lower value than did the SF [P = 0.02] or OO [P = 0.03] groups) and for urinary F2-isoprostane concentration (P = 0.04; the SF+C and OO groups had lower values than did the SF group [P = 0.02 and P = 0.03, respectively]). At T₃, cats in the SF+C group had lower TAG (P = 0.04) and urinary isoprostane (P = 0.02) concentrations than did cats in the SF group; however, the difference between pairs of groups with respect to cholesterol concentration did not achieve significance (P = 0.07). At T₃, cats in the OO group had a lower urinary F2-isoprostane concentration than did cats in the SF group, but again the difference did not achieve significance (P = 0.05).

Discussion

The purpose of the study reported here was to compare the effects of consuming 1 of 3 moderate-fat, moderate-energy diets in overweight cats. Two effects were evaluated: those of supplementation with flavonones at a nutritional dose (0.03% hesperidin, 0.017% narangin) in a moderate-fat diet with saturated animal fat as the main source of fat and those of substitution of a portion of the SF with MUFAs-rich oil. Although insulin sensitivity was not estimated, plasma lipids concentrations in all cats at T₀ were similar to values reported elsewhere for serum TAG and cholesterol concentrations in obese, insulin-resistant cats (44 ± 12 mg/dL and 140 ± 23 mg/dL, respectively).⁸

Cats’ metabolizable energy intake was lower in the SF+C-fed (7% less) and OO-fed (4% less) groups than in the SF-fed group; therefore, effects of the addition of citrus flavanones or MUFAs could not be distinguished from the effect of lower energy intake. An important limitation of the present study is that only data regarding mean daily food consumption by groups of cats were available. Therefore, variations in findings among individual cats could not be assessed, and the exact role of energy restriction in each cat could not be determined. The energy requirement of obese cats can be estimated as (130 kcal/kg of BW^0.4) and of lean cats can be estimated as (100 kcal/kg of BW^0.67), which translates to 70 kcal/kg of BW in lean cats.³⁵ The study cats had a mean food intake of 51 to 55 kcal/kg of BW, which approximately corresponds to 137 to 151 kcal/ kg of BW^0.4. However, no significant change in BW was detected in any diet group during the study.

In overweight cats, 5 months of feeding a moderatefat, moderate-energy diet with citrus flavanones supplementation (SF+C) was associated with a lower energy intake and with lower plasma lipids (cholesterol and TAG) concentrations, compared with the effects of the same diet with SF alone. The hypolipemic effects of citrus flavanones has mainly been investigated in humans and other animals with hypercholesteremia or in subjects that consumed high-cholesterol diets.^16,18–20 In our study, citrus flavanones were added to a diet containing SF as the main source of fat. No data are available about supplementation of feline diets containing other types of fat. The study cats had plasma lipids concentrations that were within the references limits of the laboratory, so possible clinical benefits of citrus flavanones supplementation remain unclear. It would be interesting to assess the effects in cats with clinical disease. For example, additional investigation is required to determine whether the lipid-lowering properties detected in our study would be clinically relevant in cats with hyperlipidemia or diseases associated with abnormal lipids concentrations.

Another study^a revealed that naringin fed at a pharmacological dose (3% of diet) limits weight gain for the same energy intake; decreases TAG synthesis and lipid accumulation in liver; decreases blood cholesterol, TAG, glucose, insulin, and very low–density lipoprotein concentration; and improves insulin sensitivity.^a Consumption of hesperidin leads to a decrease in blood and liver lipids concentrations in ovariectomized mice.¹⁶ In addition, consumption of naringin supplement appears to preserve tissue from morphological damages induced by a high-cholesterol diet.¹⁸ Liver biopsies were not performed in the present study because of ethics restraints but would have been useful to evaluate whether citrus supplementation could reduce liver lipid accumulation in cats.

Additional studies should be performed in cats to determine whether dietary supplementation with citrus flavanones could induce similar benefits in overweight cats with insulin resistance or hepatic steatosis and to improve the understanding of the mechanism of action of these flavanones. In rabbits fed a high-cholesterol diet, naringin appears to contribute to hypocholesterolemic action via downregulated activity of acyl-CoA: cholesterol acyltransferase and higher excretion of fecal sterols, compared with consumption of a control diet.¹⁸ In our study, we did not detect an increase in fecal fat excretion (data not shown). However, a limitation in intrepreting our results is the ad libitum feeding that induced lower energy intake associated with citrus flavanones supplementation, compared with intake associated with SF alone, and the lack of monitoring of individual food consumption. In obese dogs, a 43% decrease in energy intake associated with diet modification (low fat and starch and high protein and fiber content) is also associated with a decrease in blood lipids concentration.³⁸ Therefore, from the present results, one cannot make conclusions about the hypolipidemic effect of citrus flavanones alone, as consumption of this diet also resulted in lower energy intake (7% less), compared with consumption of the SF diet, although the difference in energy intake is much less substantial than in the canine study.

Five months of feeding a low-energy diet with citrus flavanones (SF+C) was associated with a lower food intake and with a lower urinary F2-isoprostane concentration than was the same duration of SF feeding. Oxidative stress may be the unifying mechanism underlying the development of comorbidities in obesity. Evidence suggests that a clustering of sources of oxidative stress exists in obesity, including hyperglycemia, hyperleptinemia, increased tissue lipid concentrations, inadequate antioxidant defenses, increased rates of free radical formation, enzymatic sources within the endothelium, and chronic inflammation.¹¹ In overweight cats, a high urinary F2-isoprostane-to-creatinine ratio has been detected.¹³ Positive effects of dietary supplementation with flavonoids have already been reported for various animal species. In dogs, oral administration of green tea polyphenols (epigallocatechin 3-gallate), grape seed polyphenols, or citrus polyphenols (naringin) can significantly decrease blood oxidation biomarkers in dogs, alter expression of genes involved in inflammation and insulin resistance, and improve lipid status.^22,39,k Results of the present study appeared to confirm antioxidant activity of citrus flavanones in cats. Antioxidant properties of bioflavonoid mixtures in cats have already been documented.⁴⁰ A study^k in lean dogs was conducted to compare the effects of food intake restriction (35% less food, without diet composition modification) or dietary naringin supplementation at a nutritional dose on antioxidant status, revealing that the status of some blood markers of oxidative stress improves after naringin supplementation, whereas food intake restriction appears to induce a higher degree of oxidative stress. On the basis of results of the present study, the decrease in urinary F2-isoprostane could have been more related to citrus flavanones supplementation than to energy restriction.

In humans, the traditional Mediterranean diet is associated with low mortality rates from cardiovascular diseases. Olive oil, the most typical component of that diet, is believed to contribute to the low mortality rates. The biological effects of olive oil may be mediated by its lipids composition (MUFA) but may also depend on or may be potentiated by its microcomponents composition.^41,42 The olive oil used in the present study was a blend between mechanically and solvent-extracted oil. Therefore, the benefits attributable to natural microcomponents were probably limited so as to evaluate mainly the effect of monounsaturated fat. In the human medical literature, the most documented effects of olive oil consumption concern improved insulin sensitivity. In rats with induced fatty liver, olive oil consumption also appears to decrease hepatic accumulation of TAG.⁴³ In our study, compared with feeding of SF, feeding of OO was associated with lower energy intake but did not induce a significant effect on plasma lipids concentrations. Insulin sensitivity measurements or liver biopsies were not performed to investigate other possible effects. In another study⁴⁴ involving obese cats, feeding a protein regimen, a high carbohydrate–high saturated fatty acid regimen, and a high-carbohydrate diet enriched with n-3 polyunsaturated fatty acids resulted in few significant effects on blood lipids values. However in that study, all diets contained more fat than was present in our study diets. The urinary F2-isoprostane concentrations in our study cats were lower (albeit insignificantly) after 5 months of feeding a diet rich in MUFAs than a diet rich in saturated fat, suggesting that consumption of the MUFAs-containing diet decreased the susceptibility of cell membranes to peroxidation, compared with results after consumption of diets high in saturated fats.

Results of the study reported here suggested that in overweight cats, lower energy intake and consumption of citrus flavanones added to a moderate-energy, moderate-fat diet with saturated fat as the main source of fat led to decreased plasma lipids concentrations (total cholesterol and TAG) and oxidative status (as indicated by urinary F2-isoprostane concentration), compared with the effects of consuming saturated fat alone. Consumption of a moderate-energy, moderate-fat diet enriched with olive oil and lower energy intake also appeared to induce a lower degree of oxidative stress (as indicated by lower urinary F2-isoprostane concentration), compared with the effects of consuming a moderate-energy diet with saturated fat as the main source of fat. However, plasma lipids concentrations remained within reference limits throughout the study, so the clinical importance of the differences detected is not known. These results require confirmation by clinical studies of cats with hyperlipidemia or diseases related to oxidative stress and abnormal blood lipids concentrations (eg, diabetes mellitus). Lower energy intake with dietary citrus flavanones supplementation or with substitution of saturated fat for monounsaturated fat could be a strategy for decreasing blood lipids concentrations or oxidative stress in overweight cats.