Novel dietary strategies can improve the outcome of weight loss programmes in obese client-owned cats

Abstract

A randomised, single-blinded, positively controlled, field trial for weight loss in obese client-owned cats was undertaken to look at novel diets and dietary strategies that could improve owner compliance and, therefore, success of feline weight loss programmes. Three dietary strategies were evaluated: strategy A used a novel dry high fibre ration; strategy B used ready-prepared portions of dry and moist food; strategy C used an existing commercial dry high fibre ration fed with a measuring cup. Cats were assessed at weeks 4, 12 and 20, and adjustments to the energy allocation made if required. Mean weight loss at 20 weeks (A: 11.0±1.3%, B: 10.9±1.2%, C: 11.9±1.7%) and mean energy allocation (A: 31.0, B: 28.5 and C: 32.2 kcal/kg of target body weight/day) were similar amongst strategies. However, owners' subjective hunger score was significantly (P<0.001) higher with strategy C than with strategies A and B. Further, owner satisfaction was lowest with strategy C, and more owners also regarded food allowance as insufficient with this strategy. Novel diets and feeding strategies may improve outcome in feline weight loss programmes.

Problems with excess body weight (BW), overweight and obesity, are now common in cats, and numerous disease predispositions exist. Although pharmaceuticals have recently received approval for the treatment of weight loss in dogs, these drugs are not licensed for cats and, therefore, conventional strategies for weight management remain the treatment of choice. Weight loss programmes have been shown to be successful in experimental trials, ^1–4 but performance in practice is less successful, with slower rates of weight loss and more marked energy restriction required to achieve success. ⁵ From the perspective of the owner, one major hurdle is the perception that energy restriction causes hunger, leading to increased scavenging activity and behaviours which the owner often interprets as begging.

Feeding behaviour is complex in cats, and is frequently misinterpreted by owners. Cats are functional feeders and eating is a solitary behaviour. ⁶ In a natural feline context the acquisition and consumption of food is not a social activity and the ability to control access to food is an important factor in reducing feline stress. Although, in the wild, cats would consider many small meals per day, ⁶ a recent survey in North American and Australian pet cats revealed that the majority were fed three or fewer meals per day. ⁷ Such practices can be inherently stressful for cats, and can lead to problems whereby humans misinterpret feline interactions at the time of feeding. ⁶ Due to the ritualised nature of the feeding process owners engage in specific social contact with their pet, through tactile and vocal interaction, during the delivery of the food. Cats respond to this interaction with social behavioural patterns, which the owner then misinterprets as signs of ‘hunger’ and a request for food. ⁶ If cats are always fed at such times, a learned association can develop between social initiating behaviours, such as rubbing and vocalising, and the delivery of food. The learning is taking place for the human as well as the cat, and the human response to feed the cat whenever it displays these social behaviours results in excessive feeding, which can be one of the factors predisposing to the development of feline obesity. Further, when dietary energy is restricted during a weight loss plan, owners will often complain that their pet displays increased levels of ‘hunger’ related behaviour and this puts increased strain on the owner-animal bond, causing owner non-compliance or complete withdrawal from the programme.

Satiety is defined as the feeling of fullness and disappearance of appetite after a meal. ⁸ Dietary factors can influence satiety, and evidence in humans suggests that some foods may be more effective than others in reducing hunger, likely due to the influence of macronutrients in the diet. ^9–11 Foods which are high in either protein, ¹² fibre, ^13,14 carbohydrates or water ^15,16 have been shown to improve satiety. However, the satiety effect of proteins is variable, because speed of digestion varies amongst different proteins, and different amino acids induce the secretion of insulin to varying degrees. Recently, a diet with increased content of both protein and fibre has been shown to improve satiety in dogs, ¹⁷ and this diet improves outcome in dogs on weight management regimes. ¹⁸ Although the triggers to feeding behaviour may be different in cats, one study has suggested that altering dietary macronutrients (eg, substituting dietary protein for fibre) can also produce a spontaneous reduction of food intake (satiety) through mechanisms other than altered diet palatability. ¹⁹ Therefore, it is possible that such diets could improve owner compliance and thus outcome for weight loss in cats.

A second problem with current weight loss regimes is the fact that significant time and commitment is required on the part of the owner, in terms of instigating a controlled feeding strategy. All such strategies require controlling food intake by measuring out daily food allocation. For dry (kibbled) diets, the most common approach is to use a measuring cup, but there can be approximately 20% variability in the amount measured by different people (unpublished observations of the authors). Using the same set of electronic weigh scales to measure out food can significantly reduce this variability, but is more time-consuming than using a measuring cup. A novel approach, that combines user-friendliness with accuracy, is to use pre-weighed portions, but such a strategy has never been tested in the field.

Based on the hypothesis that using diets with altered dietary nutrient composition and strategies promoting ease of portion allocation will improve outcome of weight loss programmes in cats, the purpose of the current study was to assess different dietary strategies for weight loss in cats.

Materials and Methods

Study outline

The study was a randomised, single-blinded, positively controlled, field trial for weight loss in client-owned cats with naturally occurring obesity. All experimental protocols complied with European Union guidelines on animal welfare and were approved by the Royal Canin committee for animal ethics and welfare. Owners gave their informed written consent to participate in the study. Cats were randomised, in blocks of three, according to the cat order at the first visit.

Recruitment of cases

A commercial company, specialised in running consumer panel trials, was used to recruit all cats, and two sources were used: some cats were recruited directly from clients on the company database, others responded to advertisements in local newspapers. The company handled all enquiries and was responsible for enrolling the cases. All cats selected were from the same urban area (Lyon, France); selected owners were then contacted by telephone, to verify the inclusion criteria, as well as the BW of their cat. The purpose of study, protocol, and potential benefits to their cat were explained, and owners' willingness to participate confirmed.

Owners that were interested in participating were then invited to present their animal for a final screening session at the Clinique des Animaux de Compagnie, École Nationale Vétérinaire de Lyon, France. For each cat, a physical examination was performed, BW was measured, and body condition score (BCS) assessed with a five-point system. A single veterinarian (TB) performed all assessments in 1 week. Inclusion and exclusion criteria used for case enrolment are listed in Table 1. Owners of cats who participated in the trial were paid 250 Euros at the 12-week point of the trial, and received all diet food free of charge.

Table 1.

Inclusion and exclusion criteria for the study.

Inclusion criteria	Exclusion criteria
Healthy	Pre-enrolment
Indoor	On a dietetic treatment
Single-cat household	Complicating treatments *
Age 2–13 years	Exclusion post-enrolment
BW: >5.0 kg (female) & >5.5 kg (male)	Food refusal and food exclusion
	Non-observance
	Other illnesses, treatment or anaesthesia

^*Complicating treatments requiring exclusion pre-treatment included corticosteroids, antibacterials, psychotropic drug, hypoglycaemic drugs, anti-hypertensive medications, and contraceptives.

Study diets

Cats were enrolled and randomised into three treatment groups, each involving a different dietary strategy. Three different diets were used, all designed for administration during weight management (Table 2). Cats on strategy A were fed a dry (kibbled) ration high in fibre, especially fibre with a high-water-binding capacity (Table 2, diet V64). Cats on strategy B, were fed a mixed ration of diet V64 and one 100 g pouch of a moist obesity diet (Table 2, diet T58). The third strategy (C) was a positive control, involving a different dry (kibbled) high fibre ration (Table 2, diet X96), although the exact characteristics were different from V64.

Table 2.

Average analysis of the three study diets.

	V64		T58		X96
% As fed	% Dry matter	% As fed	% Dry matter	% As fed	% Dry matter
Protein	34.0	36.4	7.5	48.4	35.1	37.7
Fat	9.0	9.6	2.0	12.9	9.2	9.9
Crude fibre	14.1	15.1	1.5	9.7	14.1	15.2
NFE	27.6	29.5	3.1	20.0	29.8	32.0
Starch	18.7	20.0	3.1	20.0
Dietary fibre	23.0	24.6	1.5	9.7	24.8	26.7
Ash	8.3	8.9	1.9	12.3	4.8	5.2
Moisture	7.0	–	84.0	–	7.0	–
ME (Atwater)	3274	3503.2	604	3896.8
ME (Modified Atwater)	2921	3125.5	541	3490.3	3050	3279.6

V64=Satiety Support, Royal Canin, Aimargues, France; T58=Obesity Management, Royal Canin, Aimargues, France; X96=R/D feline dry, Hill's, Topeka, USA. ME=Metabolisable energy in kcal/kg; Atwater: 4, 4, 9 kcal ME/g for protein, carbohydrates and fat, respectively; Modified Atwater: 3.5, 3.5 and 8.5 for protein, carbohydrates and fat, respectively.

Given that each treatment differed, it was not possible to blind the investigator as to the diet group. However, the owner was blinded by the fact that, although they were aware that their cat was participating in a weight loss trial, they were not aware that there were different diet groups and feeding strategies. To minimise owner bias further, plain (eg, no distinguishing features) packaging was used for all diets, and a plain feeding cup for X96 was also used, although identical in shape and size to the cup used commercially for the same diet.

Trial protocol and monitoring

The outline of the trial protocol is shown in Fig 1. On day 0, cats underwent physical examination, BW measurement, and BCS assessment (see below). Cats were then randomly allocated to the three diet groups, and instructions on amount to feed and method of feeding were given at this stage (see below). The trial was initially planned as a 12-week weight loss trial, and owners signed up on that basis. However, during the trial, a decision was made to extend the trial to 20 weeks in order to get closer to target weight; therefore, at the 12 week visit, owners were asked to attend for a further visit at 20 weeks.

Fig 1.

Time-line for the 20-week field trial for feline weight loss.

Therefore, cats were re-examined at weeks 4, 12 and 20; at each of these visits, owners received a briefing explaining the purpose of the visit as well as education on obesity, its associated problems and management strategies. All cats were then weighed and adjustments to the diet made as required (see below). A physical examination was also performed to insure that the cat had remained healthy, and a questionnaire was used to assess clinical aspects of weight loss and overall owner satisfaction with the diet (see below). All owners were then called 48–72 h after the first visit to confirm that they were happy with the trial and the advice given. In addition to the scheduled revisits, owners were called by phone on week 9 to address any problems and also to complete a questionnaire. As an additional supportive measure, a 24-h telephone hotline was available for the owners to call, if problems were encountered.

Energy restriction and method of feeding

The initial BCS, taken on day 0, was used to estimate target BW (TBW), with each 0.5 BCS unit above 3/5, corresponding to 10% excess weight. ²⁰ Cats were fed between 30 and 35 kcal/kg TBW/day, using simplified scaling criteria based upon TBW (Table 3). In strategy A, diet was offered in small neutral bags, containing 60 kcal of kibbles; in strategy B, cats were given one neutral pouch of wet diet V64, and the remaining ration in small neutral bags of T58; for strategy C, the diet was given in a plain (ie, no distinguishing features) 3 kg bag, and owners were instructed to weigh out the kibble with a plain measuring cup. This latter strategy was designed to mimic the standard approach taken in clinical practice.

Table 3.

Simplified scaling criteria for initial energy allocation in cats.

TBW (kg)	Theoretical energy allocation *		Number of units † (for 30 kcal/kg TBW)	Number of units † (for 35 kcal/kg TBW)	Final number of units †	Final energy allocation *
BCS 3.5–4.0 (30 kcal/kg TBW)	BCS 4.5–5.0 (35 kcal/kg TBW)
3.0	90	105	1.5	1.8	1.5	90
3.5	105	122.5	1.8	2.0	2	120
4.0	120	140	2.0	2.3	2	120
4.5	135	157.5	2.3	2.6	2.5	150
5.0	150	175	2.5	2.9	2.5	150
5.5	165	192.5	2.8	3.2	3	180
6.0	180	210	3.0	3.5	3	180
7.0	210	245	3.5	4.1	4	240
8.0	240	280	4.0	4.7	4	240

^*Historically, energy allocation for diets from this manufacturer, differed depending upon starting BCS, with a higher starting allocation (theoretical energy allocation) in the most obese cats (eg, BCS 4.5–5). This strategy aimed to reduce the adverse effects of energy restriction (eg, manifestation of learned hunger behaviours) at the start of the programme. However, in the current study, a simplified scaling system used where a single level of energy intake was assigned (final energy allocation) at a level within the range of 30–35 kcal/kg TBW/day.

^†A unit was defined as 60 kcal of food; for V64, this represented one 100 g pouch; for T58, this represented a single bag of 20 g, and for X96, this corresponded to the owner-measured unit of food.

At each revisit, the rate of weight loss was calculated; if the rate of weight loss was <0.5%/week, the energy allocation was decreased by 30 kcal/cat whilst for cats where weight loss was >2%/week, energy allocation was increased by 30 kcal/cat.

BW, BCS, and feline body mass index (FBMI)

BW was measured with electronic weigh scales (Presica 30000 G SCS: Dietikon, Switzerland; sensitivity 1 g, capacity 31,000 g); at each session, the scales were first calibrated to ensure precision and accuracy. BCS was estimated with a five-point condition score system. ²¹ In addition, FBMI was calculated as previously described. ²²

Assessment of wellbeing, behavioural alterations, and owner satisfaction

At each time point, owners first received a briefing explaining the purpose of the visit as well as education on obesity, its associated problems and management strategies. The owners completed a questionnaire, which assessed wellbeing, behaviour and owner satisfaction with the diet. Patient wellbeing was determined by assessing changes in activity levels, coat condition and faecal quality. For activity, owners were asked whether their cat's activity level had increased, decreased or had remained unchanged. Coat quality was graded subjectively on a five-point scale as better (scores 1 and 2), worse (scores 4 and 5) or no change (score 3). The presence of seborrhoea was assessed with a five-point score as worse (scores 1 and 2), improved (scores 4 and 5) and no change (score 3). Faecal quality was assessed with a five-point scale, as previously described. ²³ Briefly, a score of 4 would represent an optimal faecal quality, whilst lower and higher scores represent either looser or firmer stools, respectively.

Owners assessed how hungry their cat was with a subjective four-point scale (score 1: no change in hunger; scores 2–4: mild, moderate and severe increase in hunger, respectively). Other specific alterations in behaviour were also assessed as present or absent, eg, increased vocalisation, stealing food, increased owner-contact, and aggression towards the owner. Finally, owners were asked to rate satisfaction with the diet by subjectively assessing ease of use and whether portion size was adequate for their cat. Ease of use was evaluated with a 10-point scale (10 easiest, one difficult), adequacy of portion was evaluated with a simple ‘yes-no’ question.

Statistical analysis

Normally distributed results are expressed as mean±standard deviation, whilst categorical data are expressed as median (range). The primary outcomes of interest were the weight loss variables (ie, overall percentage of weight loss, rate of weight loss, and energy allocation required for weight loss), whilst secondary outcomes of interest included completion rates at 12 and 20 weeks, patient wellbeing, behavioural changes, owner satisfaction, BCS and FBMI.

Statistical analysis was performed with two computer software packages (Fizz 2.21 A, Biosystèmes, Couternon, 2007; Stats Direct version 2.6.2, Stats Direct, Altrincham, UK). A Kolmogorov Smirnov test was used to confirm that continuous data followed a normal distribution, and parametric tests were then used as appropriate, including repeated measures analysis of variance (ANOVA); the protected least significant difference (PLSD) Fisher procedure was used for post hoc analysis. Categorical data were analysed with a Kruskal–Wallis test with Dwass–Steel–Kritchlow–Fligner post hoc testing, whilst proportional data were assessed with Fisher's exact test or χ² tests, as appropriate. The level of statistical significance was set at P<0.05.

Results

Study animals

A total of 86 selected cats were identified during initial screening; two cats were recruited through the database, whilst the remaining 84 cats were recruited through the newspaper advertisements. Sixteen cats were rejected during the final screening session, 14 because their BCS was below 4/5 and two because of difficulties with handling. Therefore, a total of 70 cats fulfilled the inclusion criteria; mean BW was 7.7±1.2 kg and 6.3±1.4 kg, for males and females, respectively, whilst median BCS was 4.5 (4.0 to 5.0) and 4.0 (4.0 to 5.0), for males and females, respectively.

The 70 cats were randomised into the three dietary groups, with 22, 25 and 23 cats in groups A, B and C, respectively. Seven cats were excluded from the trial, after initial enrolment; one from group A, three from group B and three from group C. The reasons for exclusion were failure to adapt to the diet (one each for group A and group B), diet refusal (two for group B), failure of owner to comply with study requirements (two in group C), and development of concurrent illness (hepatic lipidosis; one in group C). There was no difference in completion rates at 12 weeks (A 21/22, 95%; B 23/25, 92%; C 20/23 87%; Fisher's exact test, P=0.68). A further 15 cats did not return for the final reassessment at 20 weeks of age (five from group A, four from group B, six from group C). Therefore, completion rates at 20 weeks were 16/22 (73%), 18/25 (72%) and 14/23 (61%), for diets A, B and C, respectively. There were no significant differences in completion rates amongst diets at this time point (χ² analysis, P=0.62).

Preliminary group comparisons

A total of 48 cats completed the 20-week weight loss trial, and remained healthy throughout. All statistical comparisons were made using this group of 48 cats. At day 0, the three groups were homogenous for number of animals (16, 18 and 14, respectively), BW (7.6±1.9 kg, 6.8±1.5 kg, and 7.4±1.5 kg, one-way ANOVA, P=0.41), BCS (median [range] 4.5 [4.0–5.0], 4.0 [4.0–5.0], and 4.0 [4.0–5.0], Kruskal–Wallis test, P=0.22), age (5.7±1.9, 6.3±2.2, and 6.4±3.2 years, one-way ANOVA, P=0.75) and gender distribution (group A: 10 neutered female, six neutered male; group B: seven neutered female, 11 neutered male; group C: five neutered female, nine neutered male, χ² test, P=0.26).

Outcome of weight loss

BW declined significantly at each time point, in each group of cats (Table 4, P<0.001), and the amount of weight lost was similar amongst groups (P=0.77). The mean percentage of weight lost over the course of the 20-week study was similar amongst groups (11.0±1.3%, 10.9±1.2% and 11.9±1.7%, respectively; P=0.34). Further, mean rate of weight loss did not vary significantly amongst time points for any diet (P>0.1 for all, Table 5), and there was no significant difference amongst groups for overall rate of weight loss or at any individual time point (0.53±0.21, 0.53±0.21 and 0.57±0.22% BW/week, respectively; P=0.38). Finally, mean energy allocation for the duration of the trial was not significantly different amongst groups (28.9±3.3, 31.6±4.5 and 32.2±3.0 kcal/kg of TBW/day, respectively, for strategies A, B and C; Table 5).

Table 4.

BW of cats over time with different diet strategies.

	Strategy A				Strategy B				Strategy C
	T0	T4	T12	T20	T0	T4	T12	T20	T0	T4	T12	T20
BW (kg)	7.20a±1.827	6.93b±1.729	6.59c±1.650	6.41d±1.558	6.94a±1.633	6.74b±1.514	6.38c±1.490	6.18d±1.417	7.35a±1.512	7.13b±1.512	6.78c±1.373	6.47d±1.200
FBMI	49a±12.6	46b±9.7	44b±10.3	40b±8.8	47a±9.9	42b±6.9	41b±6.6	40c±7.2	50a±11.1	46b±9.8	47ab±10.2	39c±6.1

Strategy A=diet V64 fed in pouches; strategy B=combination of diet V64 and T58, fed in pouches; strategy C=diet X96 fed with measuring cup. T0=start of diet; T4=4 weeks; T12=12 weeks; T20=20 weeks. Only the 48 cats that completed the study are included in this table. Time points with different suffixes (a, b, c and d) are significantly different from one another (P<0.05).

Table 5.

Rate of weight loss and mean energy allocation for the three strategies over time.

	Strategy A				Strategy B				Strategy C
	T0–4	T4–12	T12–20	T0–20 *	T0–4	T4–12	T12–20	T0–20 *	T0–4	T4–12	T12–20	T0–20 *
Rate of weight loss (%)	0.75±0.62	0.63±0.50	0.39±0.39	0.53±0.21	0.96±1.09	0.60±0.58	0.35±0.28	0.53±0.21	0.78±0.39	0.59±0.33	0.52±0.40	0.57±0.22
Mean energy allocation (kcal ME/kg)	33.7±1.75	31.5±4.78	30.6±5.14	28.9±3.33	30.9±2.01	29.1±3.12	27.9±3.41	31.6±4.53	33.4±1.96	32.6±3.19	30.9±2.84	32.2±3.04

Strategy A=diet V64 fed in pouches; strategy B=combination of diet V64 and T58, fed in pouches; strategy C=diet X96 fed with measuring cup. T0=start of diet; T4=4 weeks; T12=12 weeks; T20=20 weeks. Only the 48 cats that completed the study are included in this table. No significant differences were noted amongst diets or time points.

^*T0–20 represents mean values for the whole 20-week study.

BCS and body mass index

BCS decreased significantly over time for all diets (Fig 2, P<0.001); further, there was no difference, at any time point, between diets. Similarly, FBMI decreased significantly throughout the course of the study, and the change was not significantly different amongst diets (Table 4, P<0.001).

Fig 2.

Subjective begging scores, as determined by the owners, across the course of the 20-week trial. Columns represent mean values, whilst bars represent 95% confidence intervals. Scores were significantly higher for diet strategy C, than for the other two strategies (P=0.0033).

Patient wellbeing

Owners reported a subjective increase in activity with time on all diets, but the only statistical difference was for strategy A, where more cats were reported to have increased activity levels at 12 weeks than at 4 weeks (Table 6, P<0.05). Owners also reported that coat quality improved over time whilst on all diets (Table 6, Friedman's test: diet A, T0 vs T4, T12 and T20, P=0.028; diet B, T0 vs T20, P<0.001; diet C, T0 vs T4, T12 and T20, P<0.001), and there was no significant difference in coat condition amongst diets (P=0.24). In contrast, the seborrhoea score improved over time in the cats on feeding strategies A and B, but not for cats on feeding strategy C (Table 6, Friedman's test: diet A, T0 vs T4, T12 and T20, P=0.038; diet B, T0 vs T4, T12 and T20, P=0.0042; diet C, P=0.073). Finally, there was no significant difference in faecal scores amongst any of the diets. Although faecal score was significantly lower at the 4-week visit (vs weeks 12 and 20) in the cats on strategy A (Friedman's test, P<0.001), faecal scores remained above 4 throughout.

Table 6.

Changes seen by owner in activity, coat quality and faecal score with each diet strategy over time.

	Strategy A				Strategy B				Strategy C
	T0	T4	T12	T20	T0	T4	T12	T20	T0	T4	T12	T20
Activity level
Decreased	–	0/16 (0)	0/16 (0)	1/16 (6)	–	0 (0/23)	0/18 (0)	0/18 (0)	–	1/14 (7)	0/14 (0)	0/14 (0)
No change		6/16 (38)a	2/16 (11)b	3/16 (19)ab		9/18 (5)	6/18 (33)	3/18 (17)		4/14 (29)	6/14 (38)	3/14 (21)
Increased	10/16 (62)	14/16 (89)	12/16 (75)	9/18 (50)	12/18 (67)	15/18 (83)	9/14 (64)	8/14 (62)	11/14 (79)
Coat quality	3a 1–4	3b 2–5	4b 3–5	4b 2–5	3a 1–4	3.5ab 2–5	3.5ab 2–5	4b 1–5	2.5a 1–4	3b 2–5	4b 2–5	4b 2–5
Seborrhoea score	4.5a 1–5	5b 1–5	5b 4–5	5b 2–5	3.5a 1–5	4.5b 2–5	5b 2–5	5b 2–5	3.5 1–5	4.5 3–5	4.5 3–5	4 2–5
Faecal score	–	4.1±0.57a	4.5±0.52b	4.2±0.58b	–	4.4±0.50	4.4±0.51	4.3±0.57	–	4.3±0.47	4.1±0.66	4.0±0.27
Diet practicality	–	8.5±1.47	8.2±1.81	8.1±1.44x	–	8.4±1.64	8.0±1.61	8.2±1.60xy	–	6.7±2.70	6.6±3.85	6.9±2.82y

Strategy A=diet V64 fed in pouches; strategy B=combination of diet V64 and T58, fed in pouches; strategy C=diet X96 fed with measuring cup. T0=start of diet; T4=4 weeks; T12=12 weeks; T20=20 weeks. Time points with different suffixes (a, b, c) are significantly different from one another (P<0.05); diets with different suffixes (x, y, z) are significantly different from one another (P<0.05).

Behavioural alterations

Over the 20-week study, a significant diet effect was observed for owner-reported ‘hunger’ scores (Fig 2; Kruskal–Wallis test P=0.0033); post hoc analysis revealed that hunger score was significantly higher on diet strategy C compared with both of the other diets. Further, at 20 weeks, owners of cats on strategy C reported that cats vocalised more than cats on strategy B (P<0.05). A significant diet effect was observed for cats seeking owner-contact, with increased owner-contact being reported for cats fed on diet strategy C than for strategy B at the 20-week point (P<0.05). However, there was no difference, in reported frequency of aggression, amongst diets at any time point.

Owner satisfaction

Results of the owner satisfaction questionnaire are shown in Table 6. At the end of the diet, the practicality scores reported by the owner of cats on strategy A (dried food in pouches), were significantly higher than those of strategy C (measuring cup) (A: median 9, range 5–10; C: median 7, range 3–10; P=0.019; Table 5); however, there were no differences between strategies A and B and between B and C. Finally, more owners rated the volume of food given during the diet to be insufficient on diet C than on diets A and B (13/20 vs 6/20, and 8/23, respectively, χ² test, P=0.048).

Discussion

The current study has confirmed the findings of previous studies showing that weight loss can be successfully achieved in client-owned cats, although the rate of weight loss is slower than predicted from the experimental setting. ⁵ The main outcomes of weight loss did not differ amongst the three diet strategies; however, there were differences with regard to practicality and owners' perception of the behaviour of their cats. The reasons for such differences are not known, but likely relate either to a difference in diet formulation, or in feeding strategy, or both. One limitation of the study was that, as the three groups used both different diet formulations and methods of feeding (pre-prepared portion vs measuring cup), it is not possible to determine the exact reason for the differences observed. Thus, further studies, where these factors are separated, are now required to examine this in more detail.

A common complaint from owners on a weight loss programme is the fact that, when dietary intake is restricted, the learned hunger behaviours of their cat increase in frequency and intensity. Inability to cope with this behaviour is a common reason for failure of feline weight loss diets and, in the current study, a marked degree of energy restriction was required (see below). The owners of cats fed strategies A and B, reported lower hunger scores than cats on diet strategy C. Although the reason for such behavioural differences is not clear, it is noteworthy that both of these strategies included a dry diet with high-water-binding capacity fibre. A recent study has shown that voluntary food intake in cats is lower when this type of fibre is substituted for protein in the diet, with each 1% change from protein to fibre reducing intake by 1 g/day. ¹⁹ Given that palatability did not differ between the diets used in the previous work, ¹⁹ and similar diets were used in the current study, it is tempting to speculate that improved satiety (rather than palatability) may be responsible for the change in apparent hunger in this study. This conclusion would be supported by the fact that vocalisation and owner-seeking behaviour was also reduced on these strategies.

Owners preferred the diet strategies which involved feeding of pre-determined portions of food, rather than using a measuring cup. Given a strategy involving pre-weighed portions is simpler for owners, it is likely that compliance will be improved when cats undergo a long weight management programme. Such a strategy should also avoid the potential for inaccuracies to arise from using measuring cups (unpublished observations of the authors).

Owners reported a number of benefits on patient wellbeing, most notably increases in activity (for diet strategy A), coat quality and reduction in the presence of seborrhoea. The changes in activity levels are not surprising, and would support subjective impressions of changes in activity during weight loss in cats. Although the reason for improved coat quality is not clear, one explanation would be that weight loss enables improved ability to groom. However, as cats on the different strategies lost a similar amount of weight, grooming changes alone are unlikely to explain the differences observed between the diets for seborrhoea score. An alternative possibility would be that differences in diet composition, eg, essential fatty acids (EFAs) could account for these changes. However, against this, all diets were sufficiently EFA supplemented so that recommended allowance would be comfortably met even at the level of energy intake provided. Therefore, the exact reasons for the differences in seborrhoea score are not clear and require further investigation in the future.

A novel approach was used to recruit privately owned cats for a clinical study. In this respect, a commercial recruitment company and a newspaper advertising campaign were used, and a financial reward was given when cats completed 12 weeks of the trial. To our knowledge, this is the first time that such an approach has been used in a clinical study. The main benefits were the fact that a large number of subjects were rapidly identified, in a small geographical area, and this allowed all cats to be started at the same time (eg, within the period of a week). Not all cats identified in the original screening process, were appropriate for enrolment in the study, and the most common reason for exclusion was the fact that the cat was not obese (eg, BCS<4/5). The fact that it was relatively common for some owners to over-estimate the degree of obesity in their pet cat, contrasts with previous studies suggesting that owners may underestimate their cats true body condition. ²⁴ The method of recruitment may also have accounted for the fact that compliance during the first 12 weeks of the study was excellent with ≥87% of cats remaining in the study; thereafter, the drop-out after this point was greater. The relatively high level of attrition between 12 and 20 weeks most likely reflects the fact that owners had originally only committed to the first 12 weeks, and that there was no further financial incentive for continuing. Nonetheless, most owners did decide to continue, suggesting that, once committed, most owners were keen to see the weight programme through.

The rate of weight loss (approximately 0.53–0.57% BW/week) and mean energy allocation (28.5–32.2 kcal/kg TBW/day) during the study was similar amongst groups. Thus rate of weight loss was slower and the degree of energy restriction greater than previous studies, both in research settings ^1–4 and in other clinical studies. ⁵ The reason why such modest rates of weight loss were achieved is not clear, but could relate to the relatively infrequent re-examinations, eg, four times in 20 weeks, which is less frequent than other clinical studies, eg, every 14–28 days. ⁵ Most notably, there were 8-week intervals between the final three visits (4, 12 and 20 weeks). Thus, it is likely that the rate of weight loss would have been greater if more visits had been undertaken, as more frequent adjustments could have been made. The degree of energy restriction is again of note and, as with other studies, ⁵ indicated that cats in a clinical setting require lower energy allocations to achieve an adequate degree of weight loss. In fact, this level of energy restriction required on diet strategy C was greater than the manufacturers would typically recommend for the diet used, and may have contributed to the problems encountered with satiety.

In summary, the work presented in this study has shown that using diet strategies that involve both altering dietary macronutrient composition and method of feeding, can improve weight loss outcomes. The main advantages include a reduction in owner-perceived begging behaviour (appetite) and improved ease of use, and such benefits are likely to improve the outcome of weight loss strategies for cats. However, further studies are required to determine whether these benefits result from altering macronutrient composition (eg, protein vs fibre content) or altering the method of feeding (eg, pre-prepared portions vs measuring cup).