Short-term changes in dietary fat levels and starch sources affect weight management, glucose and lipid metabolism, and gut microbiota in adult cats

Abstract

A 2 × 2 factorial randomized design was utilized to investigate the effects of fat level (8% or 16% fat on a fed basis) and starch source (pea starch or corn starch) on body weight, glycolipid metabolism, hematology, and fecal microbiota in cats. The study lasted for 28 d and included a low fat and pea starch diet (LFPS), a high fat and pea starch diet, a low fat and corn starch diet, and a high fat and corn starch diet. In this study, hematological analysis showed that all cats were healthy. The apparent total tract digestibility of gross energy, crude protein, and crude fat was above 85% in the four diets. After 28 d, cats fed the high fat diets (HF) gained an average of 50 g more than those fed the low fat diets (LF). The hematological results showed that the HF diets increased the body inflammation in cats, while the LFPS group improved the glucolipid metabolism. The levels of glucose and insulin were lower in cats fed the LF diets than those in cats fed the HF diets (P < 0.05). Meanwhile, compared with the LF, the concentrations of total cholesterol, triglyceride, and high-density lipoprotein cholesterol in serum were greater in the cats fed the HF diets (P < 0.05). Additionally, both fat level and starch source influenced the fecal microbiota, with the relative abundance of beneficial bacteria, such as Blautia being significantly greater in the LFPS group than in the other three groups (P < 0.05). Reducing energy density and using pea starch in foods are both valuable design additions to aid in the management of weight control and improve gut health in cats. This study highlights the importance of fat level and starch in weight management in cats.

Reducing fat levels and using amylose starch in commercial cat diets was a valuable nutrition strategy for glycemic control and gut microbial regulation in domestic cats.

Introduction

Obesity has become a prevalent health issue among domestic cats (German, 2006; Cave et al., 2012; Öhlund et al., 2018). The causes of feline obesity are multifaceted, primarily involving factors such as diet, feeding practices, age, and gender (Wall et al., 2019). Feline obesity is associated with alterations in gut microbiota, inflammation, and the development of diabetes (Rand, 2013; Fischer et al., 2017; Martins et al., 2023). Commercial cat food typically contains high fat (HF) levels, but high content of fat in diet has been identified as a risk factor for overweight cats (Nguyen et al., 2004; Backus et al., 2007). According to the recommendations by the Association of American Feed Control Officials (AAFCO), the minimum requirement for fat content in complete cat food is 9% on a dry matter basis to meet the nutritional needs of adult cats (AAFCO, 2014). Many commercial cat foods contain HF during the maintenance phase, resulting in a positive energy balance for cats. Reducing dietary fat levels and restricting calorie density are effective strategies for managing obesity in obese cats (Linder and Freeman, 2010). Meanwhile, previous studies have shown that overweight impair the restoration of islet β cell functions, leading to diabetes in cats (Hoenig, 2012; Clark and Hoenig, 2021). However, few studies have investigated the effects of the reduction of dietary energy density on healthy cats, with most studies focusing on the effects of lower dietary fat levels on obese cats.

Furthermore, it is noteworthy that commercial cat food is produced using an extrusion process to bind, shape and improve palatability (Baller et al., 2021). This process requires expandable raw materials, typically represented by starch with good bulking properties. Domestic cats can digest carbohydrates (Baller et al., 2021), and a study reported that when foods with similar palatability, healthy adult cats were able to respond to metabolic demands in macronutrient choices and chose foods with carbohydrates accounting for 43% of calories (Hall et al., 2018). The effects of different types of starch on blood sugar changes vary widely. Therefore, selecting expandable ingredients with a controlled glycemic response is a viable option for reconciling the conflict between health concerns and manufacturing process. Corn starch (CS) is a common ingredient in cat food, but it may not be conducive to optimum blood sugar control in cats due to its fast digestion rate and high digestibility (de-Oliveira et al., 2008). In contrast, pea starch (PS) contains slowly digested amylose and which has been proven to enhance glucose metabolism regulation in rat model (Boaventura et al., 2023). Most studies on the effect of starch source were conducted in obese or diabetic cats (Appleton et al., 2004; Hoenig et al., 2007b), and related research in healthy cats is also limited.

Gut microbes are closely related to the health and disease of cats, and dietary management is the most important factor affecting the gut microbial composition (Wernimont et al., 2020). While there is limited research on the effects of HF foods on gut microbiota in cats, extensive research in humans and rodents has documented that the HF diets could lead to gut microbial dysbiosis and subsequent disease (Kovatcheva-Datchary et al., 2019; Tong et al., 2021). Compared to dietary fats, carbohydrates are generally complex and highly heterogeneous in structure, and thus degraded and fermented by microorganisms in the hindgut to both affect gut health through produced metabolites and yield basic energy salvage (Keenan et al., 2015; Zhang et al., 2023).

To our knowledge, few studies combined the effects of fat level and starch source to enhance nutrition strategy for obesity prevention in healthy cats. Therefore, the objective of our study was to explore the effects of joint alterations in dietary fat level and starch source on weight management, glucolipid metabolism, and gut microbiota in healthy adult cats.

Materials and Methods

Animals and experimental design

All protocols for the study were approved by the China Agricultural University Institutional Animal Care and Use Committee (approval number: AW21402202-1-7, Beijing, China). The study involved forty spayed or neutered domestic adult cats (50% male, 50% female) with a mean body weight of 3.45 ± 0.85 kg and a mean age of 3.07 ± 0.36 yr. We scored each cat on a body condition score based on Laflamme’s 9-point scale (Laflamme, 1997), and all the cats were scored as 6. According to the hematology results, all cats are healthy.

The research employed a 2 × 2 factorial randomized block design to study the effects of fat level (8% or 16% poultry fat on a fed basis) and starch source (PS or CS). Forty cats were randomly assigned to four treatment groups, with 10 replicates in each group, with half males and half females. The dietary treatments consisted of a low fat and pea starch diet (LFPS), a high fat and pea starch diet (HFPS), a low fat and corn starch diet (LFCS), and a high fat and corn starch diet (HFCS). All diets are produced through extrusion expansion, with an extrusion temperature ranging from 130 to 140 °C and an extrusion chamber pressure of 50 to 80 MPa. Diets were analyzed for feed ingredients and chemical composition, with two replicate samples for each index (Table 1), and all diets met the nutritional requirements recommended by the NRC (2006) (National Research Council., 2006). Before the experiment, all cats were not given antibiotics for 3 mo and were fed the same basic diet. The basic diet ingredients and nutrient analysis are shown in Supplementary Table S1. The formal experiment lasted for 28 d, which included a 4-d period of fecal collection on days 21 to 24.

Table 1.

Diet ingredients and nutrient analyses

Variable	Treatments1
	LFPS	HFPS	LFCS	HFCS
Ingredient, %
Chicken meal	48.3	40.3	48.3	40.3
Pea protein	5.00	5.00	5.00	5.00
Fish meal	7.00	7.00	7.00	7.00
Poultry fat	8.00	16.0	8.00	16.0
Corn starch	—	—	30.0	30.0
Pea starch	30.0	30.0	—	—
Salt	0.45	0.45	0.45	0.45
Taurine	0.15	0.15	0.15	0.15
Choline chloride	0.40	0.40	0.40	0.40
Magnesium sulfate	0.25	0.25	0.25	0.25
Vitamin premix2	0.30	0.30	0.30	0.30
Mineral premix3	0.20	0.20	0.20	0.20
Total	100	100	100	100
Nutrient level
Moisture, %	5.20	5.70	5.90	6.00
Gross energy, kcal/kg	4825	5135	4825	5302
Crude fiber, %	1.50	1.41	1.00	1.91
Crude protein, %	47.3	40.1	46.9	41.7
Crude fat, %	10.4	18.4	11.1	17.8
Ash, %	9.78	8.23	8.48	8.36
N-free extract4	25.8	26.2	26.6	24.2
Starch, %	26.9	26.5	26.8	26.3
Amylose/amylopectin ratio	0.41	0.39	0.24	0.22
Calculated ME5 (kcal/kg)	3443	3883	3617	3856

¹Values are the means (n = 2). HFCS, high fat and corn starch diet; HFPS, high fat and pea starch diet; LFCS, low fat and corn starch diet; LFPS, low fat and pea starch diet.

²Vitamin premixes provided the following per kilogram of feed: vitamin A (15,000 IU), vitamin B₁ (30 mg), vitamin B₂ (28 mg), vitamin B₃ (110 mg), vitamin B₅ (85 mg), vitamin B₆ (12 mg), vitamin B₁₂ (0.19 mg), vitamin D₃ (15 IU), vitamin E (75,300 IU).

³Mineral premixes provided the following per kilogram of feed: Ca (CaI₂) 20 mg, Co (CoSO₄) 0.10 mg, Cu (CuSO₄) 3 mg, Fe (FeSO₄) 50 mg, I (CaI₂) 40 mg, Mn (MnSO₄) 18 mg, Na (Na₂SeO₃) 0.05 mg, Zn (ZnSO₄) 38 mg, Se (Na₂SeO₃) 260 mg.

⁴N-free extract, % = 100 – (crude protein + crude fat + crude fiber + moisture + ash).

⁵Calculated with modified Atwater equation: Metabolizable energy (ME) (kcal/kg) = 10 × (3.5 × crude protein + 8.5 × crude fat + 3.5 × N-free extract).

Feeding and management

In the whole trial, cats were provided ad libitum access to water and food, and were housed in cages individually (1.5 m × 1.5 m × 2.0 m). The humidity and temperature were controlled at 50% to 60% and 23 to 26 °C, respectively. On days 0 and 28, cats were weighed. All cats were scored for their mental status on days 7, 14, 21, and 28, and an average mental status score (AMSC) was calculated using an 8-point scale (blinking = 1, snoring = 1, walking = 1, tail wagging = 1, pole climbing = 2, front paw extension = 2). Greater scores were associated with better mental status. The amount of feed consumed was weighed daily to determine the average daily feed intake (ADFI).

Sample collection

On days 0 and 28, after a 12-h fasting period, blood samples of 2 mL were collected from 40 cats via the forelimb vein and transferred to Vacutainer tubes. One blood sample was immediately frozen at −80 °C for hematology analysis, while the other was centrifuged at 3,000 × g for 30 min at 4 °C to obtain the serum, which was then stored at −80 °C for future analysis. Fecal samples were collected from 40 cats using the total collection method during days 21 to 24. The feces were collected and weighed daily, and then frozen at −80 °C. All collected feces were mixed and ground with a high-speed multi-function grinder. On day 28, fecal samples were immediately collected after defecation from six randomly selected cats in each group. These samples were snap-frozen in liquid nitrogen and stored at −80 °C for later analysis of fecal microbial composition and short-chain fatty acid (SCFA) concentration.

Chemical composition

According to the Official Methods of Analysis of AOAC International (Horwitz, 2010), feed ingredients were analyzed in terms of moisture, crude protein, crude fiber, crude fat, and ash. Gross energy was determined by an automatic adiabatic oxygen bomb calorimeter (Parr 1281 Automatic Energy Analyzer, Moline, IL, USA). The total starch (A148-1-1) and amylose (A152-1-1) were detected by commercial chemical kits according to the kit instructions (Nanjing Jiancheng Institute of Biological Engineering, Nanjing, China). The amylopectin was calculated by using the after equation:

$$amylopectin=totalstarchamylose$$ (1)

The value for N-free extract was calculated as following:

$$N-freeextract=100\left(\begin{array}{l}crudeprotein+crudefat\\ +crudefiber+moisture+ash\end{array}\right)$$ (2)

All chemical compositions were analyzed in duplicate. Metabolizable energy (ME) was calculated as follows:

$$ME=\left(\begin{array}{l}3.5\times crudeprotein+8.5\times crudefat\\ +3.5\times N-freeextract\end{array}\right)\times 10$$ (3)

Nutrient digestibility and energy metabolism

Nutrient digestibility and energy metabolism tests were performed on days 21 to 24. All feces of the cats were collected and feed intake was recorded during this time period. The fecal samples were thawed at 4 °C and then dried at 65 °C for 72 h. Before analysis, feed and fecal samples were ground through a 40 mesh (425 µm) sieve. The moisture, crude protein, crude fat, and ash of the manure were analyzed according to the same methods as for the diets. The apparent total tract digestibility (ATTD) of nutrients and energy metabolism were calculated by using the after equations:

$$ATTD\text{ = (}nutrientintakefecalnutrientoutput\text{) /}nutrientintake\times \text{100 }$$ (4)

$$grossenergyintake\text{ = }feedintake\times grossenergyofthediet$$ (5)

$$fecalenergy\text{ = }fecalweight\times fecalenergy$$ (6)

$$apparentdigestibleenergy\text{ = }grossenergyintakefecalenerg$$ (7)

Serum parameters measurement

The levels of total cholesterol (TC; A111-1), triglycerides (TG; A110-1-1), and high-density lipoprotein cholesterol (HDL-C; A112-1-1) in the serum were measured using assay kits from Nanjing Jiancheng Bioengineering Institute, according to the kit instructions. Insulin was measured using the enzyme-linked immunosorbent assay kit from Mercodia (Uppsala, Sweden; 10-1233-01). Glucose was measured using a colorimetric glucose oxidase method (Thompson, 1966). Glucagon-like peptide 1 (GLP-1; orb409852), glucagon-like peptide 2 (GLP-2; orb409503), G protein-coupled receptor 41 (GPR41; orb439513), G protein-coupled receptor 43 (GPR43; orb756784); leptin (orb390949) and, peptide YY (PYY; orb441860) were measured using assay kits from Biorbyt (Cambridge, CB4 0WY, UK) according to the instructions.

Hematology analysis

Routine blood parameters were measured by the automatic animal blood cell analyzer (Mindray BC2800, Shenzhen, China).

Determination of fecal SCFAs

Fecal SCFA contents were determined by gas chromatography according to our previous study (Wu et al., 2019).

Fecal microbiota analysis

To prepare the samples for sequencing, total microbial genomic DNA was extracted from fecal samples using the Fecal Genomic DNA Extraction Kit (Tiangen, Beijing, China). The V3 to V4 region of the 16S rRNA gene was then amplified using universal bacterial primers (338F and 806R) and sequenced on the Illumina MiSeq platform. Raw sequencing data were subsequently quality-controlled using fastp (v0.19.6), with paired-end reads being merged using FLASH (v1.2.7) and denoised using DADA2. Taxonomic analysis was carried out using QIIME 2 (v2022.2) with the Silva138/16s_bacteria database at a confidence threshold of 0.7. Alpha diversity was analyzed using R-3.3.1 (stat) software, with Kruskal–Wallis rank-sum test and Tukey–Kramer post hoc analysis for inter-group differences. Group differences were tested by principal co-ordinates analysis (PCoA) with analysis of similarities at the genus level using the Bray–Curtis distance. Linear discriminant analysis (LDA) effect size (LEfSe) was used at different taxonomic levels on phylum and genus with R-3.3.1 software, all-against-all (most strict) comparison, and an LDA threshold of three. All the P-values were subjected to false discovery rate correction. The raw sequencing data has been deposited in the NCBI Database under SRA accession number PRJNA910297.

Statistical analysis

The study performed was a two-factor analysis of variance (ANOVA) model. The model accounted for the effects of fat levels, starch sources, and their interaction. Statistical analyses were performed using the Generalized Linear Model procedure of IBM SPSS Statistics 25. The significance of interactions and main effects was obtained from the ANOVA table. P < 0.05 was considered statistically significant, and 0.05 ≤ P < 0.10 was ­considered a tendency. When the interaction was significant (P < 0.05), Tukey’s post hoc comparison was performed among the four treatment groups. The data were shown as means and pooled SEM of the four groups.

Results

Body characteristics and nutrient digestibility

The statistical analysis indicated that the interaction and main effects of fat level and starch source on the measured parameters, including body weight, ADFI, and AMSC, were not significant among the groups (P ≥ 0.05; Table 2). Furthermore, the AMSC values observed in cats were found to be within the normal reference range during the 28-d feeding period.

Table 2.

Body characteristics of cats fed different diets

Variable	Treatments1				SEM	P-value
	LFPS	HFPS	LFCS	HFCS		Fat	Starch	Interaction
Body weight
Day 0, kg	3.53	3.28	3.55	2.99	0.13	0.12	0.61	0.54
Day 28, kg	3.54	3.34	3.62	3.09	0.12	0.15	0.74	0.51
ADFI, g/d	77.5	78.2	85.0	78.3	3.34	0.50	0.76	0.44
AMSC	3.16	4.44	4.00	3.40	0.32	0.73	0.73	0.10

¹Values are means of each treatment group with their pooled standard error of the means (SEM); n = 10/treatments. The interaction values were analyzed by Tukey’s post hoc comparisons. Significantly different values had different letters in the same row (P < 0.05). ADFI, average daily feed intake; AMSC, average mental status score; HFCS, high fat and corn starch; HFPS, high fat and pea starch; LFCS, low fat and corn starch; LFPS, low fat and pea starch.

Nutrient digestibility and energy metabolism

The results of nutrient digestibility and energy metabolism are presented in Table 3. During the energy metabolism test, cats fed the HF diets exhibited greater levels of apparent digestible energy compared to cats fed the low fat (LF) diets (P < 0.05). Additionally, there was a trend of greater gross energy intake in cats fed the HF diets compared to the LF diets (P < 0.10). In the ATTD experiment, all experimental groups demonstrated digestibility rates above 85% for gross energy, crude protein, and crude fat. Cats fed the HF diets exhibited greater levels of apparent digestible energy compared to cats fed the LF diets. However, the interaction and main effects of the ATTD of fat, protein, and ash were not statistically significant (P ≥ 0.05).

Table 3.

Energy metabolism and apparent total tract digestibility of nutrients in cats

	Treatments1				SEM	P value
	LFPS	HFPS	LFCS	HFCS		Fat	Starch	Interaction
Energy metabolism
Gross energy intake, kcal	1257	1482	1386	1619	63.4	0.07	0.29	0.97
Apparent digestible energy, kcal	1114	1374	1228	1536	62.5	0.02	0.25	0.84
Digestibility, %
Gross energy	88.6	92.7	88.6	94.8	0.01	<0.01	0.36	0.38
Crude protein	91.2	89.9	90.0	91.2	0.60	1.00	0.95	0.32
Crude fat	89.9	95.8	95.7	96.3	0.59	0.18	0.18	0.26

¹Values are means of each treatment group with their pooled standard error of the mean (SEM); n = 10/treatment. The interaction values were analyzed by Tukey’s post hoc comparisons. Significantly different values had different letters in the same row (P < 0.05). HFCS, high fat and corn starch; HFPS, high fat and pea starch; LFCS, low fat and corn starch; LFPS, low fat and pea starch.

Hematology analysis

The hematological parameters of the cats were analyzed before the study and the results were shown in Supplementary Table S2. All values were within the normal range and were not different among the four groups (P ≥ 0.05). After the 28-d feeding experiment, the results were presented in Supplementary Table S3. All hematological parameters were within the normal reference range, indicating that cats were in normal health status. However, the interaction between fat level and starch source had a significant effect on red blood cell count and mean platelet volume in cats (P < 0.05). Specifically, cats in the HFCS group had greater contents of red blood cell counts and mean platelet volumes than those in the other three groups (P < 0.05). In relation to the erythrocyte system, cats fed the HF diets had greater levels of ­hemoglobin, hematocrit, mean corpuscular volume, and mean corpuscular hemoglobin than cats fed the LF diets (P < 0.05). Additionally, the levels of white blood cells, lymphocyte percentage, monocytes, neutrophils, eosinophils, platelets, and basophils were greater in cats fed the HF diets than those in cats fed the LF diets (P < 0.05). In contrast, cats consuming the PS diets had lower white blood cell counts, platelets and mean platelet volumes than cats consuming the CS diets (P < 0.05).

Serum glucose lipid metabolism and inflammation

The serum biomarkers indicating glycolipid metabolism in cats were shown in Table 4. The interactive effect of fat level and starch source was observed on TC levels (P < 0.05). Compared to cats fed the LF diets, cats fed the HF diets had lower levels of glucose and insulin (P < 0.05). In contrast, the serum levels of TG, TC, and HDL-C were greater in the HF group than those in the LF group (P < 0.05). Furthermore, cats consuming the CS diets had greater levels of glucose, insulin, TG, TC, and HDL-C than cats consuming the PS diets (P < 0.05). The serum hormones of cats were shown in Table 4. The serum concentrations of PYY and leptin were lower in the LF diets than those in the HF diets (P < 0.05). Similarly, GLP-1 and GLP-2 were greater in cats fed the HF diets than those in cats fed the LF diets (P < 0.05). Additionally, cats fed the LF diets had lower levels of GPR41 and GPR43 compared to cats fed the HF diets (P < 0.05). Furthermore, the concentrations of PYY, leptin, GLP-1, and GLP-2 were lower in cats fed the PS diets than those in cats fed the CS diets (P < 0.05). In contrast, the serum levels of GPR41 and GPR43 were greater in cats consuming the CS diets than cats consuming the PS diets (P < 0.05).

Table 4.

Serum glucolipid metabolism and hormones of cats on day 28

Variable2	Treatment1				SEM	P-value
	LFPS	HFPS	LFCS	HFCS		Fat	Starch	Interaction
TG, mmol/L	4.41	6.23	5.95	7.70	0.21	<0.01	<0.01	0.88
TC, mmol/L	5.55c	7.64b	7.36b	8.85a	0.20	<0.01	<0.01	0.04
HDL-C, μmol/L	900	1310	1259	1607	44.2	<0.01	<0.01	0.45
Glucose, mmol/L	30.8	38.4	40.8	46.8	1.03	<0.01	<0.01	0.34
Insulin, mU/L	40.9	52.6	53.6	61.7	1.28	<0.01	<0.01	0.13
PYY, pmol/L	8.95	12.3	12.3	16.1	0.45	<0.01	<0.01	0.61
Leptin, ng/mL	23.7	31.5	31.2	41.0	1.05	<0.01	<0.01	0.22
GLP-1, pmol/L	2.41	4.83	4.22	6.20	0.32	<0.01	<0.01	0.20
GLP-2, pmol/L	3.74	7.26	8.69	13.4	1.05	<0.01	<0.01	0.27
GPR41, ng/mL	4.42	7.35	7.27	9.59	0.24	<0.01	<0.01	0.25
GPR43, ng/mL	23.7	31.5	31.2	41.0	0.60	<0.01	<0.01	0.22

¹Values are means of each treatment group with their pooled standard error of the mean (SEM); n = 10/treatment. Means were analyzed by a two-factor ANOVA model, and the interaction values were analyzed by Tukey’s post hoc comparisons. Significantly different values had different letters in the same row (P < 0.05). GLP-1, glucagon-like peptide 1; GLP-2, glucagon-like peptide 2; GPR41, G protein-coupled receptor 41; GPR43, G protein-coupled receptor 43; HDL-C, high-density lipoprotein cholesterol; HFCS, high fat and corn starch; HFPS, high fat and pea starch; LFCS, low fat and corn starch; LFPS, low fat and pea starch; PYY, peptide YY; TC, total cholesterol; TG, triglyceride.

Fecal SCFA concentrations

The effects of fat level and starch source on fecal SCFA concentrations of cats were shown in Supplementary Table S4. Fat level and starch source did not contribute to significant interactions for acetic acid, propionic acid, butyric acid, isobutyric acid, valeric acid, isovaleric acid, and total SCFA (P ≥ 0.05).

Fecal microbiota

The effects of starch source and fat level on fecal microbial composition in cats were shown in Figure 1. Our findings indicated that cats fed the HF diets had greater abundance of Peptoclostridium than cats fed the LF diets. Conversely, the abundance of Negativibacillus, Eubacterium_brachy_group, Fusobacterium, Odoribacter, norank_f__norank_o_Clostridia_UCG-014, and Parabacteroides were lower in cats consuming the HF diets than those in cats consuming the LF diets (P < 0.05). Additionally, cats fed the PS diets had greater abundance of Ruminococcus_torques_group, norank_f__Eubacterium_coprostanoligenes_group and Blautia than cats fed the CS diets (P < 0.05). The fecal microbiota composition among the four treatments were displayed in Figure 2. The Shannon and Simpson indexes of cats fed the LFPS and HFCS diets were greater than those of cats fed the HFCS diet (P ≥ 0.05; Figure 2A and B). Furthermore, PCoA analysis revealed a significant difference in microbial composition among the four treatments (P < 0.05; Figure 2C). At the phylum level, Firmicutes, Bacteroidota and Actinobacteriota were the most abundant bacteria (Figure 2D). Based on the family level in the LEfSe analysis (Figure 2E), the LFPS group was enriched in differential bacteria with an LDA score > 4.0, including Eiryipeiotrichaoea, norank_o_Clostridia_UCG-014, Christensenellaceae, and Monoglobaceae. The genus-level LEfSe bar (Figure 2F) showed that among the four treatment groups, the LFPS group showed greater abundance of Blautia, Ruminococcus, norank_f_orank_Clostridia_UCG-014, and norank_f_Eggerthellaceae than the other groups (P < 0.05), while the HFCS group showed greater abundance of Megasphaera and Bifidobacterium than the other groups (P < 0.05).

Figure 1.

Bar plot of fecal microbial differences under two factors (fat level and starch source). (A) Genus-level microbial differences: Wilcoxon bar plot with different fat levels. (B) Genus-level microbial differences: Wilcoxon bar plot with different starch sources. LF, composed of LFCS and LFPS; HF, composed of HFCS and HFPS; PS, composed of LFPS and HFPS; CS, composed of LFCS and HFCS. Each treatment included 12 replicates. Significant differences were presented as * P < 0.05, ** P < 0.01, and *** P < 0.001. CS, corn starch; HF, high fat; HFCS, high fat and corn starch; HFPS, high fat and pea starch; LF, low fat; LFCS, low fat and corn starch; LFPS, low fat and pea starch; PS, pea starch.

Figure 2.

Description of the fecal microbial composition in cats. (A) Shannon index. (B) Simpson index. (C) Principal coordinate analysis (PCoA) plots of bacterial communities at the genus level. (D) Community bar plot analysis at the phylum level. (E, F) Linear discriminant analysis (LDA) effect size was used to analyze the fecal bacteria with significant abundance at the family and genus levels among the different groups; LDA scores > 3. Each treatment included 6 replicates. LFPS, low fat and pea starch; HFPS, high fat and pea starch; LFCS, low fat and corn starch; HFCS, high fat and corn starch.

Discussion

Obesity is a common health concern for pet cats and is gaining increasing attention (German, 2006; Cave et al., 2012; Öhlund et al., 2018). Rapid weight gain in cats is partly due to the excessive energy intake (Nguyen et al., 2004; Laflammme, 2010; Gooding et al., 2014). Commercial cat foods often contain elevated fat levels. Meanwhile, previous studies in cats have shown that the ATTDs of fat were above 90% when the dietary fat levels based on dry matter ranged from 10% to 30% (Butowski et al., 2019; Von Schaumburg et al., 2021). In our study, cats fed diets with 8% or 16% fat both had more than 89% ATTDs of fat. Our results also revealed that cats consuming the HF diets had a greater calorie intake and significantly greater apparent digestible energy compared to cats on the LF diets. Meanwhile, dietary fat has the lowest thermogenic effect and is highly efficient in providing energy to the body. And body fat accumulation resulted from the high level of fat as well as the efficient digestion and utilization of nutrients. Our study also found that cats fed the HF diets gained an average of 50 g more weight compared to cats fed the LF diets over the course of the 28-d experiment. This finding further supported the notion that a high content of dietary fat can increase the risk of obesity in cats.

As expected, increases in serum TC and TG concentrations were observed after 4 wk of feeding the HF diets of cats. These results were consistent with previous studies conducted on obesity cats (Hoenig et al., 2006). Numerous studies in the past have also found greater concentrations of TG in cats after feeding the HF diets (Thiess et al., 2004; Keller et al., 2017). The high digestibility of dietary fat with greater dietary fat intake might explain the elevated serum TC and TG levels. Interestingly, it has been reported that high serum levels of TC and TG, causing lipid accumulation in muscles and insulin resistance, leading to a weakened glucose metabolism (Nishii et al., 2012). In this study, the greater TG, TC, and HDL-C levels in the CS group than in the PS group could be partly attributed to the rapid digestion of CS, resulting in the reversal of the glucose-fatty acid cycle, and the oversupply of glucose which can inhibit lipid oxidation (Kraegen et al., 2001), ultimately inducing lipid storage.

Dietary fat and carbohydrate both influence cats’ glucose and lipid metabolism, as well as hormone responses (de-Oliveira et al., 2008; Coradini et al., 2013). A compensatory increase in insulin secretion is required to maintain normal glucose tolerance, but the prolonged and excessive demand for extra insulin is postulated to eventually lead to pancreatic β-cell “exhaustion” and to contribute to the development of feline diabetes (Hoenig, 2012; Clark and Hoenig, 2021). According to the carnivore connection theory proposed that diets that stimulate greater long-term insulin secretion contribute to the excessive demand for insulin (Miller and Colagiuri, 1994). The results showed that the HF and CS diets significantly increased the levels of blood glucose and insulin. These results indicate that the HF and CS diets may stimulate an increase in blood glucose levels and result in compensatory insulin secretion.

Then, we analyzed the relevant hormones involved in cat glucose metabolism. Although there is limited research on these hormones in cats, immunopositive cells for GLP-1 and PYY have also been found in the gastrointestinal tract of cats (Gilor et al., 2013). GLP-1 stimulates insulin secretion, inhibits gastric empty, and promotes satiety (Ellingsgaard et al., 2011). Cats exhibit a similar postprandial increase in GLP-1 response (Model et al., 2022). Additionally, when compared to amino acids and glucose, lipids have been found to be more effective in stimulating GLP-1 secretion in cats (Gilor et al., 2011). Leptin has been shown to inhibit insulin secretion from the pancreatic β cells, the increase of leptin may contribute to the impaired insulin response to glucose (Appleton et al., 2002). Moreover, both PYY and leptin are anorexigenic hormones that can increase satiety (Roth et al., 2018; Pizarroso et al., 2021). Studies have shown that cats exhibit a greater plasma PYY response after consuming a large amount of food, which may reduce hunger (Camara et al., 2020). Additionally, plasma leptin levels in cats have been positively correlated with body weight (Appleton et al., 2002; Hoenig et al., 2007a). Our results also found significantly greater GLP-1, PYY, and leptin levels on the HF diets than on the LF diets. In our study, we observed that greater dietary fat intake resulted in elevated serum leptin levels; however, all leptin values remained within the reference ranges established for normal-weight and healthy cats (Appleton et al., 2000). These findings suggest that the HF diets can increase satiety and reduce insulin sensitivity in cats. Additionally, cats fed with CS exhibited greater levels of GLP-1, PYY, and leptin compared to those fed with PS. Amylopectin starch has a greater surface area in contact with digestive enzymes and is broken down into glucose faster than amylose starch, causing blood sugar to rise faster (Behall et al., 1988). The differential content of resistant starch in the two diets may be a contributing factor to these findings. Overall, our research suggests that HF and PS diets may cause mild insulin resistance in healthy cats.

Subsequently, we conducted a hematology analysis. The results of the hematology showed that all hematological parameters were within the normal range, indicating the absence of any significant changes that could suggest a pathological condition in healthy cats. However, our study also observed significant changes in the ematological parameters of cats under different diets. The hemoglobin of cats with the intake of the HF diets was significantly greater than that of the LF diets. In addition, the mean corpuscular volume of cats in the CS group was significantly greater than that in the PS group. Notably, a study has found that the blood concentration caused by severe water loss in the plasma of obese individuals leads to a general increase in hemoglobin, hematocrit, and mean corpuscular volume in obese cats (Martins et al., 2023). The HF diets often triggers chronic inflammation in the body, leading to an increase in platelets, neutrophils, lymphocytes, and monocytes (Tanner et al., 2007; ­Furuncuoglu et al., 2016). An increase in body weight in cats has been shown to significantly elevate oxidative damage to lipids, proteins, and DNA (Tanner et al., 2007). Furthermore, research has revealed that cats exposed to HF hyperlipidemia for 10 d experience inflammatory responses (Zini et al., 2010). Notably, from a dietary perspective, although the current research does not directly establish diets as the direct cause of inflammation in cats, factors contributing to inflammatory responses in cats include the quality of dietary fat (Park et al., 2011) and high energy intake (Nguyen et al., 2004). In addition, one study has revealed that body weight change in cats offered energy in excess was dependent on dietary macronutrient profile (Allaway et al., 2018). These diet factors can increase inflammation in cats and offer intriguing clues for understanding the mechanism of inflammation. Therefore, the impact of macronutrient intake and composition should be continued attention, and further investigations will contribute to revealing the association between nutrients and body inflammation. Our results showed that the main effects of fat level on the above indicators were significant, as the serum white blood cells, lymphocyte, monocyte, neutrophil, eosinophil, and basophil counts in cats with the intake of HF diets were significantly greater than those with the intake of LF diets. These results suggest that the HF diets may induce a state of stress and elevate the risk of inflammation in cats.

Previous studies have demonstrated the impact of gut microbiota on maintaining overall feline health and its role in a variety of disease conditions (Mondo et al., 2019). Dietary factors play a crucial role in driving gut microbiota composition and function, to a greater extent than genetic factors (Schoeler and Caesar, 2019; Wernimont et al., 2020). The predominant phyla of felines were Firmicutes, Bacteroidetes, Actinobacteria, Proteobacteria, and Fusobacteria in previous studies (Ritchie et al., 2008), which were similarly found here. In this finding, PCoA analysis showed that different diets induced significant changes in microbial community structure. In the four treatment groups, the diversity and richness of microbiota were the greater when cats consumed LFPS diet. PS has a greater concentration of amylose compared to CS. The amylose will enter the large intestine for microbial fermentation and utilization, improving the intestinal microbial structure (Li et al., 2023).

We further analyzed the differences in specific gut microbes in this experiment. Blautia is an important core genus of bacteria in the mammalian gut, and its abundance was negatively correlated with markers of obesity-related metabolic disorders (Benítez-Páez et al., 2020). A large-scale human survey revealed that Blautia is the only gut microbiota significantly and negatively associated with visceral fat (Ozato et al., 2019). Decreased Blautia is associated with insulin resistance in obese individuals, while increased Blautia helps reduce inflammation associated with obesity-related complications, which may be related to the ability of Blautia to produce acetic acid (Wang et al., 2020; Yan et al., 2022), and reduce obesity by modulating GPR41 and GPR43 (Kimura et al., 2011, 2013). Elevated acetic acid produced by intestinal microbiota has been shown to have a significant protective effect on intestinal inflammation in mice (Yan et al., 2022). In this study, there was a tendency to increase the content of acetic acid in the LFPS group, which may be related to the increased Blautia abundance caused by PS. Meanwhile, a study identified that neutered in obese cats had increased abundances of Blautia (Fischer et al., 2017). Furthermore, in this study, the relative abundance of Christensenellaceae family, which has negative correlation with obesity and inflammatory bowel disease in humans (He et al., 2018), was greater in the LF group. In addition, at the genus level, our study observed that the abundance of Ruminococcus and Lactobacillus was greater in cats feeding the LFPS diet than those in cats feeding the other three diets. Research has shown that the HF diets could result in lipid and glucose metabolism disorders with greater levels of insulin-resistance indices and visceral fat rate, which was accompanied by a significant decrease of Ruminococcus, a prevalent gut microbe allowing disassembly of complex polysaccharides (Miao et al., 2021). Lactobacillus is regarded as a type of “friendly” bacteria, and which increased abundance was associated with improved insulin resistance and obesity-induced cognitive impairment (Maifeld et al., 2021). Additionally, in the present study, we also found that cats in the LFPS group had greater abundance of Marvinbryantia than cats in the other groups. It has been reported that mice with chronic colitis had dysbiosis of the gut microbiota with decrease in some typical bacterial taxa like Marvinbryantia, while the normalization gut microbial dysbiosis and the attenuation of chronic colitis were accompanied by the greater Marvinbryantia abundance (Zhai et al., 2019). Therefore, these above results suggested that the LFPS diet had a beneficial modulation on the gut microbiota structure of cats.

In conclusion, this study demonstrates that feeding cats a HF diets increases their energy intake and leads to sight weight gain. It also results in short-term increases in blood glucose and insulin secretion. Additionally, the study reveals that both fat levels and starch sources influence the composition of the adult cat fecal microbiota, with a tendency towards altered levels of acetate in SCFAs. In summary, our study suggests that reducing fat content in commercial cat food may be more favorable for weight management in cats. Additionally, supplementing with PS can have a positive impact on the composition of the gut microbiota.

Glossary

Abbreviations:

AAFCO

Association of American Feed Control Official

ADFI

average daily feed intake

AMSC

average mental status score

ATTD

apparent total tract digestibility

CS

corn starch

GLP-1

glucagon-like peptide 1

GLP-2

glucagon-like peptide 2

GPCRs

G protein-coupled receptors

GPR41

G protein-coupled receptor 41

GPR43

G protein-coupled receptor 43

HDL-C

high-density lipoprotein cholesterol

HF

high fat

HFCS

high fat and corn starch

HFPS

high fat and pea starch

LDA

linear discriminant analysis

LEfSe

effect size

LF

low fat

LFCS

low fat and corn starch

LFPS

low fat and pea starch

ME

metabolizable energy

TC

total cholesterol

TG

triglyceride

PS

pea starch

PYY

peptide YY

SCFAs

short-chain fatty acids

SEM

standard error of mean

Author Contributions

Yi Wu (conceptualization, resources, methodology, writing – review & editing), Ruixia Mo (project administration, conceptualization, methodology, visualization, writing – original draft), Mingrui Zhang (investigation, supervision, writing – review & editing) Haotian Wang (data curation, software, investigation), Tianyi Liu (software, supervision, data curation), and Gang Zhang (formal analysis, investigation, validation). All authors have read and approved the final manuscript

Conflict of interest statement

The authors have no conflicts of interests to disclose.