Evaluation of a pet-separating automatic feeder and high-frequency meal feeding for weight loss in multi-cat households

Abstract

Objectives

The objectives of this study were two-fold: (1) to determine if a feeder system providing individualized food access and automatic dispensing of meals makes weight-loss programs in multi-cat homes easier for cat owners; and (2) to assess if feeding six meals daily reduces food-seeking behavior and increases calorie requirements vs two meals per day.

Methods

Adult cats with body condition scores (BCSs) of 7/9 or greater underwent a weight-loss plan for up to 6 months using a traditional feeding bowl with two meals per day (B2) or an automatic feeder with two meals (A2) or six meals (A6) per day. Cats were weighed at weeks 0, 2, 4, 6, 10, 14, 18, 22 and 26 and food intake was adjusted to maintain weight-loss rates between 0.5% and 1.5% body weight per week. Owner questionnaires were completed at each study visit.

Results

Twenty-three cats were included in the study. Cats fed from the automatic feeders were more likely to reach ideal body condition (P = 0.006), with 83.2% of cats in A6, 40% of cats in A2 and no cats in B2 achieving a BCS of ⩽6/9. Owners of cats fed from automatic feeders reported the weight-loss plan was easier vs owners of cats fed from bowls (P = 0.01). Owners using the automatic feeders also reported fewer incidences of certain food-seeking behaviors. Neither percent body weight lost per week nor calorie intake per kg differed between groups.

Conclusions and relevance

Feeding six instead of two meals daily did not result in faster rates of weight loss or the need to increase calorie intake to maintain appropriate rates of weight loss. However, use of an automatic feeder that provided separated, portioned meal feeding made the weight-loss program easier, reduced food-seeking behaviors and improved the overall success of the weight-loss plan.

Introduction

Obesity is one of the most common diseases of domestic cats, with approximately 25–67% of cats being overweight or obese.^1–4 Feline weight-loss programs can be challenging. In a study evaluating a therapeutic weight- loss diet, only 14% of cats reached their ideal weight at the end of 6 months. ⁵ In a long-term study evaluating regain in cats that completed a weight-loss program, almost half gained weight and over 25% regained more than 50% of the weight lost. ⁶ Results cited in a research abstract revealed two major challenges to feline weight-loss programs: owner adherence and begging behaviors. In that study of 32 overweight cats enrolled in a 2-year weight-loss program, 22% were dismissed early owing to these issues. ⁷

Eating and digesting food requires energy, and researchers have hypothesized that eating frequent meals can aid weight loss. ⁸ A meta-analysis of research in humans compared 15 studies evaluating meal frequency with respect to changes in fat mass and lean mass. ⁸ While the analysis did find a negative correlation between meal frequency and adiposity, the results appeared skewed by a single study, and most studies evaluated found no difference. Domestic cats evolved as solitary hunters that sustained themselves on small prey, with approximately 8–12 frequent meal feedings per day. ⁹ Allowing pet cats unlimited access to food for grazing throughout the day predisposes to overeating and obesity.^10,11 Cat owners commonly provide only 1–2 meals daily. ¹² When infrequent meals are combined with energy restriction, unwanted attention-seeking or stress-displacement behaviors related to hunger may result. In a recent small study of 12 human patients, there was no difference in self-reported hunger in patients who ate an isocaloric diet divided into three vs eight meals per day. ¹³ The impact of small, frequent meals on hunger behaviors in cats requires further study. One aim of the current project was to determine if owners detect more begging behaviors when cats are fed two vs six meals per day.

While weight-loss programs can be challenging for cats and their owners, the addition of multiple cats with different dietary needs within the same household presents even more obstacles. Owners often struggle to separate normal-weight cats that prefer to graze throughout the day from overweight cats needing energy restriction. With advances in technology, there are now automated feeders available that will dispense predetermined portions of food and control unwanted pet access. Some feeders scan subcutaneous microchips to allow pet access to a feeder, while others use tags or collars to grant or deny food access. In a recent study that employed several modes of technology (digital scales, automated feeders, activity monitors and pet treat cameras) during a weight-loss program, cat owners reported home scales and feeders to be valuable tools. ¹⁴ The primary goal of this study was to determine if owner-reported begging behaviors of cats undergoing a weight-loss plan were improved when cats were fed from an automated feeder instead of having the owner dispense food into a bowl. This study also evaluated the effect of meal frequency on begging behavior and calorie requirements for weight loss.

Materials and methods

This study was approved by the University of Tennessee’s Institutional Care and Use Committee (UTK-IACUC 2424-0316). Study participants were recruited through campus email announcements, and local referring veterinarians were notified by email, facsimile and website advertisements. Cat owners provided written informed consent prior to enrollment.To meet the enrollment criteria, client-owned cats had to be >1 year of age, have body condition scores (BCSs) ⩾7/9, ¹⁵ and be free from medical conditions that would affect appetite or metabolic rate or preclude use of the study diet (eg, diabetes mellitus, hyperthyroidism, chronic kidney disease or chronic enteropathy). Cats also had to live indoors, share a home with at least one other cat residing primarily indoors and be willing to exclusively eat a dry food diet. At week 0, cats underwent a physical examination, complete blood count, serum chemistry analysis, thyroid hormone analysis and urinalysis to screen for medical conditions that would exclude enrollment. Using a permuted block randomization method, cats were randomly assigned to one of three groups: twice-daily feeding from a traditional food bowl (B2); twice-daily feeding from an automatic feeder (A2); or six daily feedings from an automatic feeder (A6; PortionPro Rx; Vet Innovations). The automatic feeder uses radiofrequency identification technology embedded in a collar tag to uniquely identify individual animals to the feeder. The feeder would recognize the collar tag to allow or deny access to the food dish through placement of a plastic shield over the bowl. To prevent food sharing, all household cats in groups A2 and A6 were provided with automatic feeders. No automated feeders were used in the B2 households, and owners were told to mitigate food sharing through traditional methods such as room separation during meal feedings. Feeding amounts (1/8 cup increments) and meal frequency (two or six times daily) were programmed into the feeder by a trained veterinary technician at the first visit.

A single investigator blinded to treatment assignment performed the initial BCS and body fat index (BFI) assessment. ¹⁶ BFI and BCS were used to estimate body fat percentage, and ideal body weight (IBW) was calculated with the following formula: 100 – %fat = %lean; [current body weight (BW) × (%lean/100)]/0.8 = IBW at 20% body fat. All cats were fed the same therapeutic weight-loss diet (Hill’s Prescription Diet Feline Metabolic Dry Food; Hill’s Pet Nutrition) and calorie recommendations were based on IBW using the equation [70 × (IBW_kg^0.75)] × 0.8. ¹⁷ Cats were weighed at weeks 0, 2, 4, 6, 10, 14, 18, 22 and 26, and food intake was adjusted to maintain weight loss rates between 0.5% and 1.5% BW per week. The automated feeder would allow only 1/8 cup incremental adjustments (approximately 37 kcal using the study diet). To keep adjustments consistent across groups, both bowl and automated feeder group calorie changes were based on 1/8 cup increments. Online owner questionnaires were completed at each study visit (Figure 1). Though unvalidated, questions covered changes in overall activity, hunger and quality of life. Owners recorded changes in behavior associated with hunger and stress such as inappropriate elimination, aggression toward humans or other animals, increased vocalization and/or begging for food. Owners were also asked to rate satisfaction and ease of the weight-loss program.

Figure 1.

Online questionnaire completed by cat owners at weeks 2, 4, 6, 10, 14, 18, 22 and 26

Statistical methods

ANOVA was used to test for group differences in starting BW and estimated percent body fat, and total percent weight lost during the study. Kruskal–Wallis tests were used to test for group differences in mean age and starting BCS. Pearson’s χ² test was used to determine if sex or number of cats achieving estimated IBW differed by group. Repeated-measures mixed-model ANOVAs with an AR(1) covariance structure were used to test for group and time differences in outcomes (percent BW loss per week and calorie intake per kg). Responses to the owner survey questions (changes in overall activity, hunger, quality of life, and satisfaction and ease of the weight-loss program) were averaged over the treatment period. Food-seeking behaviors were recoded as a cumulative sum. Owing to the non-normal distribution of the responses to the survey questions, groups were compared using Kruskal–Wallis tests. Statistical analyses were performed with SPSS 28 (IBM). For all analyses, a P value of <0.05 was considered significant.

Results

Twenty-four cats were initially screened, but one cat was euthanized at week 2 owing to an unrelated medical condition. Therefore, data from 23 cats were included in the study. Cats were randomly assigned to group B2 (n = 6), A2 (n = 10) or A6 (n = 7). One cat was initially assigned to A6 but refused to use the automatic feeder and was switched after the week 2 visit to the B2 group. Week 2 data from that cat were omitted from the analysis and all other data were analyzed with the B2 group. Final group distribution was B2 (n = 7), A2 (n = 10) and A6 (n = 6). Median age (3 years) and BCS (9/9) did not differ by group (Table 1). Cats in the B2 group had numerically higher estimated percent body fat (50.0%) vs A2 (42.7%) and A6 (40.5%); however, a significant difference was not detected, presumably owing to low sample size (P = 0.06). The starting BW of cats in B2 was higher (P = 0.035), but this group also had a higher proportion of males (86%) than the other two groups (50%). All cats had normal muscle condition scores throughout the study. Owners reported that cats consumed all of their prescribed food amounts unless stated otherwise. Cats were considered to have reached the target level of weight loss if they achieved a BCS of 4–6/9. This range was used to account for the subjectivity and variability of estimating body fat from BCS and BFI. ¹⁶

Table 1.

Baseline descriptions of cats by group

	Group	Mean ± SD	Median	n (%)
BW(kg)	A2	6.82 ± 1.38	6.92
	A6	6.31 ± 0.60	6.26
	B2	8.68 ± 2.41	7.82
BCS	A2	8.40 ± 0.84	9
	A6	8.17 ± 0.98	8.5
	B2	8.57 ± 0.79	9
Estimated %BF	A2	42.7 ± 4.67	44
	A6	40.5 ± 5.2	39
	B2	50 ± 8.9	50
Estimated IBW (kg)	A2	4.84 ± 0.71	4.86
	A6	4.67 ± 0.26	4.62
	B2	5.36 ± 0.75	4.99
Age (years)	A2	4.70 ± 3.43	4
	A6	3.83 ± 2.86	3
	B2	2.83 ± 1.17	3
Castrated male	A2			5 (50.0)
	A6			3 (50.0)
	B2			6 (85.7)
Spayed female	A2			5 (50.0)
	A6			3 (50.0)
	B2			1 (14.3)

BW = body weight; BCS = body condition score; BF = body fat; IBW = ideal body weight

Three cats did not complete 26 weeks of the study or reach a BCS of ⩽6/9 (two in B2 and one in A2). Owner-elected withdrawal (one in B2), owner relocation (one in A2) and inability to separate cats during feeding (one in B2) were reasons for incompletion. Therefore, five cats in B2, nine in A2 and six in A6 completed the study. Groups differed in achievement of ideal body condition (P = 0.006) with 83.2% of cats in A6, 40% of cats in A2 and no cats in B2 reaching a BCS of ⩽6/9 (Figure 2). The total percent BW lost was numerically lower in group B2 (9.7%) vs A2 (15.2%) and A6 (17.1%), but this was not statistically significant and post-hoc analysis determined that a much larger sample size would be needed to detect a difference. Mean ± SD percent BW lost per week was 0.68 ± 0.086 and did not differ by group or week. Similarly, mean calorie intake per kg (B2: 32.6 ± 5.3; A2: 37.6 ± 8.1; A6: 38.0 ± 4.2) did not differ between groups or weeks. Three cats in the study (one from each group) needed calories restricted below their initial prescribed amount. Only one 1/8 cup reduction was needed for these cats to maintain a weight loss rate between 0.5% and 1.5% BW per week. Six cats (one in B2, three in A2 and two in A6) in the study required a single 1/8 cup increase in calories to maintain their prescribed rate of weight loss.

Figure 2.

Percentage of cats in each group reaching a body condition score of 4–6/9

Owing to the lack of variability across time points, responses for questions 1–13 in the owner survey were averaged across time points. No significant differences were found between groups for questions 1–10 regarding perceived cat hunger and cat activities. Satisfaction did not differ between groups (P = 0.309). Group differences were found with question 12 (‘How easy or difficult was this weight loss plan?’; P = 0.011). Pairwise comparisons found that the score for B2 (4.0) was statistically significantly lower than A2 (4.7; P = 0.025) and A6 (5.0; P = 0.004), with no differences between the two A groups (P = 0.352).

Food-seeking behaviors (question 13) were compared between the three groups using Kruskal–Wallis tests due to the non-normal distribution of the data. Group differences were found with ‘stood at the food bowl and begged’ (P = 0.030), ‘Follow you prior to feeding him/her’ (P = 0.005) and ‘paw/bat at you prior to feeding him/her’ (P = 0.024). Pairwise comparisons found that the B2 group exhibited these behaviors more than A2 and A6, with no differences found between the automatic feeder groups (Figure 3).

Figure 3.

Food-seeking behaviors from question 13 in the owner questionnaire were compared between the three groups using Kruskal–Wallis tests due to the non-normal distribution of the data

Discussion

Weight-loss programs in multi-cat households are challenging. Cats in negative energy balance often exhibit unwanted food-seeking behaviors, and separating animals for meals can be burdensome. Developing tools and strategies to make weight-loss plans easier is necessary to combat the feline obesity epidemic. This study demonstrates that a feeder system that provided individualized food access and automated meal portioning made the weight-loss program easier for cat owners. The automated feeder also reduced owner-reported incidences of certain food-seeking behaviors. This suggests that cats began to dissociate the pet owner with meal feeding. However, the use of automated feeders did not appear to improve overall satiety or reduce all food-seeking behaviors.

The automatic feeder in this study provided two conveniences for the cat owners: the ability to separate cats during meals and accurate portioning of food. Whether owners felt the automatic feeder weight-loss program was easier due to cat separation or food portioning was not evaluated. Therefore, the results of this study may not be directly transferable to feeder systems that only separate cats or only provide food portioning. Most cats adjusted well to wearing the collars and tags and using the automated feeder system, with only one cat refusing to use the automated feeder.

Cats eating from the automatic feeder were more likely to reach their ideal body condition than cats eating from the bowls. No cats in the bowl group were able to achieve their ideal body condition, despite being assigned a similar calorie intake to the automated feeder cats. Separation of household cats in the B2 group may have been more difficult, allowing food stealing to occur. Although not statistically significantly different, cats in the B2 group also had a higher mean estimated percent body fat at the start of the study, which would increase the time needed to reach ideal body condition and likely contributed to the lower success rate in this group.

Feeding small, frequent meals has been postulated as a method to improve satiety and increase energy expenditure through the thermic effect of digestion. This study compared cats eating six vs two meals per day of the same food in an isocaloric fashion. Cats eating six meals per day (A6) had similar rates of weight loss and calorie intake as cats eating two meals per day (A2). Owners in the two automatic feeder groups also reported similar instances of food-seeking behaviors. These results suggest increased meal frequency has little clinical effect on weight loss in cats. However, the sample size of the study was small and more cats may be needed to fully determine the impact of frequent meals.

Conclusions

Feeding six instead of two meals daily did not result in faster rates of weight loss or the need to increase calorie intake to maintain appropriate rates of weight loss. More frequent meals also did not reduce the frequency of hunger-related behaviors. Cat owners using automatic feeders that portion meals and separate animals described the weight-loss program as easier than those using a traditional feeding bowl. Cats eating from bowls were also less likely to achieve ideal body condition. The use of these types of feeders may improve the success of weight-loss programs in multi-cat households.