Equations used to develop commercial dog food feeding guidelines and Canadian owner feeding practices in 2018

Abstract

Feeding guidelines on commercial dog food packages provide a suggested starting point for food provision for dogs. The equations used to develop commercial dog food feeding guidelines and the owner feeding practices surveyed were examined in this study. The equations used to develop feeding guidelines and the energy density calculation method (Traditional Atwater or Modified Atwater equation) were investigated for 200 dry dog foods sold in Canada. Not all energy densities of products were calculated using the modified Atwater equation, a requirement if claiming compliance with the Association of American Feed Control Officials (AAFCO). Commercial feeding guidelines provide conservative estimates of pet dog energy needs. A survey of dog owners’ feeding practices was conducted, with 739 responses analyzed. Respondents generally took appropriate action to manage the weight of their dogs through both exercise and dietary management. Further investigation should explore how owners may be successfully managing their dogs’ weight without veterinary supervision.

Introduction

Dog obesity is a growing concern, with some estimates placing over 50% of the pet dog population in different geographical areas in this weight category (1). Maintaining a healthy body weight throughout life is key to preventing adverse health outcomes associated with obesity, such as diabetes and osteoarthritis (1). Quality of life and longevity may also be improved by preventing or treating obesity (2,3). Sufficient food provision is important to ensure lack of malnutrition and nutrient deficiencies resulting from underfeeding or a nutrient dilute diet formulation. Underweight dogs may also have a shorter survival time from the time of diagnosis of certain conditions compared to ideal weight or overweight dogs; lean body mass, however, improves survival time (4–6). The type and quantity of food that owners feed plays a role in a dog’s body weight throughout life. Understanding the owner’s influence on the maintenance of healthy body weight in dogs is critical for identifying what potential avenues exist to prevent and treat obesity; however, few investigations exist for dogs in Canada.

Pet foods sold in Canada must comply with the Consumer Packaging and Labelling Act and the Competition Act, and all nutritional adequacy claims should be substantiated. The Guide for the Labelling and Advertising of Pet Foods further details specific labelling standards, but these are not law and can be adopted voluntarily. The Association of American Feed Control Officials (AAFCO) uses the modified Atwater factors to calculate caloric density of pet foods; therefore, any pet food products with an AAFCO claim should also use this algorithm. The European Pet Food Industry Federation (FEDIAF), in contrast, uses the Atwater equation. There are other well-accepted methods of predicting energy density of pet foods (7–11).

In addition to the accurate prediction of caloric density, accurate prediction of energy requirements is used to predict food allowances. The algorithms used to predict energy requirements and then calculate food allowances, however, are not regulated by AAFCO and companies can use different algorithms depending on what the diet is intended for. A meta-analysis conducted by Birmingham et al (12) further highlighted the variation in maintenance requirements of adult dogs. Most pet dog owners feed their pets a commercial diet (13,14), so it is critical to develop sufficiently accurate systems for determining the energy density of food as well as the appropriate recommended intakes to maintain a healthy body weight. The United States National Research Council (11) presented many possible equations to predict the energy requirements of dogs based on factors such as age, breed size, housing, and activity.

The first objective of the study herein was to determine the methods used to calculate metabolizable energy density and feeding guidelines for Canadian commercial dry dog foods. We expected that all products would be calculating metabolizable energy density using the modified Atwater factors. Equations used to develop feeding guidelines for weight control and senior/low activity diets were expected to use lower energy requirement algorithms, while foods intended for active dogs would use moderate to high energy requirement algorithms. Finally, we predicted that energy equations used to develop feeding guidelines differ based on dog size and activity level but are not specific to particular breeds. The second objective was to explore the feeding practices of dog owners and gain insight into owners’ overall attitudes about weight management in dogs. We hypothesized that Canadian owners would be engaging in practices that are potential contributors to dog obesity.

Materials and methods

Pet food product analysis

A review of dog food brands was conducted, with purposive selection of Canadian manufactured foods that had an AAFCO nutritional adequacy statement on the bag and at least 1 product per brand that met the inclusion criteria. Manufacturer and brand websites were used to collect information on the metabolizable energy content of foods, guaranteed analysis, and feeding guidelines for each product. Only extruded kibble and baked dry foods were considered. Products designated for a particular life stage or purpose (e.g., senior, toy breed, all life stages) were included. Weight control diets intended to maintain animals prone to weight gain were included, but diets explicitly for weight loss purposes were excluded. Therapeutic diets or veterinary exclusive foods were excluded to represent the range of products available to consumers that do not require veterinary oversight. Puppy diets were excluded.

The guaranteed analysis of each product was used to calculate the metabolizable energy content using the modified Atwater factors (15):

$$\text{Metabolizable\hspace{0.17em}energy}=[(3.5*\text{crude\hspace{0.17em}protein})+(8.5*\text{crude\hspace{0.17em}fat})+(3.5*\text{nitrogen-free\hspace{0.17em}extract})]*10)$$

and traditional Atwater factors (16):

$$\text{Metabolizable\hspace{0.17em}energy}=[(4*\text{crude\hspace{0.17em}protein})+(9*\text{crude\hspace{0.17em}fat})+(4*\text{nitrogen-free\hspace{0.17em}extract})]*10)$$

equations. Ash content of all diets was assumed to be 7% unless specified in the guaranteed analysis. The metabolizable energy stated on the product packaging was compared to both calculated metabolizable energy values. The percent error between the calculated and stated metabolizable energy was determined. The calculation method that yielded the smaller percent error was the method used by the manufacturer. If neither estimate yielded a percent error ≤ 8% the method of metabolizable energy calculation was considered as unknown. This criterion was chosen to avoid assigning a calculation method when there was a large difference between the stated and calculated metabolizable energy, while recognizing that the guaranteed analysis does not represent true analytical values but a range to accommodate for nutrient density differences, and will result in some error when used as the basis of the calculation.

To determine the energy requirement equations that were used to calculate feeding recommendations for each product, the lowest and highest body weight (kg) in the feeding guidelines and the corresponding recommended food intake (cups) were used. The metabolizable energy density reported on each product (kcal/cup) was multiplied by the cups of food recommended to obtain the recommended total dietary metabolizable energy allowance. This recommended metabolizable energy intake (kcal/day) was compared to the metabolizable energy intakes suggested by the US NRC (11) metabolizable energy requirement equations (Table 1). The NRC equation that provided the closest value to the product’s labeled value, determined by a percent error calculation, was considered the equation that was used. Some products present guidelines in columns for varying levels of activity of the pet (e.g., less/more, low, average/normal, high activity). Where this was observed, the guidelines for average or normal activity were used. If no equation provided a value ≤ 10% of that of the product it was considered an unknown method of calculation [could not be determined (CD)].

Table 1.

Prevalence and frequency of use of 7 predictive equations for maintenance energy requirement as reported by the US NRC (11) to develop feeding guidelines for low and high body weight in 200 Canadian dry dog food products with a total of 584 recommended intakes.

		Usage for development of feeding guidelines
		Body weight
Equation label	US NRC equation (kcal/kg BW0.75)	Low n (%)	High n (%)
E1	70	31 (10.7)	37 (12.7)
E2	95	60 (20.5)	84 (28.8)
E3	105	43 (14.7)	90 (30.8)
E4	130	55 (18.8)	21 (7.2)
E5	140	17 (5.8)	7 (2.4)
E6	180	15 (5.1)	1 (0.3)
E7	200	6 (2.1)	1 (0.3)
CDa	—	65 (22.3)	51 (17.5)

^aCD — could not be determined. BW — body weight.

Pet owner survey

This study was approved by the University of Guelph Research Ethics Board (REB #18-03-015). The survey included 36 multiple choice questions which were grouped by owner demographics and household makeup (9 questions); exercise routines of the dog(s) (3 questions); feeding practices (16 questions); and owner perceptions of body weight and self selection of their dog’s body condition score (BCS) based on pictographs (17) (7 questions). The first survey question was the consent form, which allowed agreeing participants to access the questions.

Respondents were recruited between April 28 and July 8, 2018, inclusively. The online link to the survey was posted to social media and circulated by e-mail. Sharing of the link was encouraged. At no point was personal information collected, and respondents could withdraw from the survey or choose to not answer questions at any time. Respondents were required to be at least 18 y old and the owner of at least 1 dog.

Statistical analysis

Descriptive statistics for each question were reported as number of respondents per question, as well as actual counts and percentage of responses for each answer within each question. Data were analyzed with SAS software (version 9.4; SAS Institute, Cary, North Carolina, USA). The PROC FREQ command was used to analyze each question. A Chi-squared (X²) test was used to compare predicted and actual values; predicted values were considered equal across all available options. Significance was declared at P < 0.05.

Results

Pet food product analysis

Product selection

Two hundred products from 44 brands were included. All carried an AAFCO statement that claimed the product meets the established nutrient concentrations for the product’s target AAFCO profile. Products were categorized based on the purpose specified on the product packaging as follows: maintenance or all life stages with no breed size specified (68 products), senior or weight management (47 products), toy or small breed (36 products), large or giant breed (35 products), performance (8 products), or breed specific (6 products).

Metabolizable energy

Four foods were excluded from the metabolizable energy analysis because the necessary information was not available online. Reported metabolizable energy content ranged from 2973 to 4820 kcal/kg. The modified Atwater equation was the most frequently used method of calculating metabolizable energy content and was used for 59.2% (116/196) of products. The traditional Atwater equation was used for 38.8% (76/196) of products, rather than the AAFCO endorsed modified Atwater equation. The method of metabolizable energy density determination was undetermined for 2% (4/196) of products, due to the percent error for both equations being > 8% or the percent error for both equations being equal.

Feeding guidelines

All 200 products were included in the analysis of feeding guidelines. Some products provided feeding guidelines with a range of intakes for a single weight. This resulted in a total of 584 recommended intakes that were compared to the US NRC equation recommended intakes. Of the 7 equations, the most frequently used US NRC equation was E2 (24.7%; 144/584), followed closely by E3 (22.8%; 133/584); 19.9% (116/584) of the equations used were undeterminable. For low body weights specifically, the highest percentage of equations (22.3%; 65/292) was undeterminable, and the most frequently used equations were E2 (20.5%; 60/292) followed by E4 (18.8%; 55/292). For high body weights, the most frequently used equations were E3 (30.8%; 90/292) and E2 (28.8%; 84/292) (Table 1). Figure 1 shows the percentage use of each equation within each product category.

Figure 1.

Percent usage of 7 predictive equations for maintenance energy requirements as reported by the US NRC (11) for development of feeding guidelines within each product category identified in a review of 200 Canadian dry dog food products. E — Equation used; CD — cannot determine the equation used.

Pet owner survey

The results presented are based on data from the Canadian respondents (n = 739). The remaining respondents were non-Canadian (total N = 1032). Response totals vary per question as not all questions applied to all respondents, and partially completed surveys were analyzed.

Owner and pet demographics (Table 2)

Table 2.

Overview of demographics reported by Canadian dog owners participating in an online survey about dog owner feeding practices.

Question	Frequency	Prevalence	X2	P-value
Q3. How many dogs do you own? (n = 739)
A. 1	408	55.21	95.53	< 0.0001
B. 2	200	27.06
C. 3	73	9.88
D. 4	36	4.87
E. 5+	22	2.98
Q4. With which gender do you predominantly identify? (n = 738)
A. Male	116	15.72	175.33	< 0.0001
B. Female	601	81.44
C. Non-binary or gender non-conforming	14	1.9
D. Prefer not to say	7	0.95
Q7. What is your highest level of education? (n = 739)
A. No schooling received	0	0.00	69.83	< 0.0001
B. Some elementary school but no high school	0	0.00
C. Some high school but no diploma	13	1.76
D. High school diploma or equivalent	64	8.66
E. Vocational training	11	1.49
F. Some college but no degree	63	8.53
G. Some university but no degree	51	6.90
H. College diploma	192	25.98
I. Bachelor’s degree	215	29.09
J. Master’s degree	82	11.1
K. Doctoral degree	48	6.50
Q8. How many adults (18 years of age or over) reside in your household? (n = 737)
A. 1	89	12.08	88.10	< 0.0001
B. 2	483	65.54
C. 3	100	13.57
D. 4+	65	8.82
Q9. How many children (17 years of age or under) reside in your household? (n = 736)
A. 0	545	74.05	187.30	< 0.0001
B. 1	90	12.23
C. 2	75	10.19
D. 3	23	3.13
E. 4+	3	0.41

X² — Chi-squared.

Most respondents identified as female (601/738; P < 0.0001; Q4) and owned 1 dog (408/739; P < 0.0001; Q3). Respondents were generally well-educated, with over 70% of respondents holding a college diploma or higher degree (537/739; P < 0.0001; Q7). Respondents primarily resided in households with 2 adults (483/737; P < 0.0001; Q8) and no children (545/736; P < 0.0001; Q9).

Exercise routines (Table 3)

Table 3.

Overview of the exercise habits of dogs reported by Canadian dog owners participating in an online survey about dog owner feeding practices.

Question	Frequency	Prevalence	X2	P-value
Q11. For an average day, what would be the dominant exercise that your dog(s) perform(s)? If you have multiple dogs that perform different amounts of exercise, choose the level of exercise performed by the most active dog. (n = 729)
A. Less than 30 minutes of walking	90	12.35	64.56	< 0.0001
B. 30–60 minutes of walking	195	26.75
C. More than 60 minutes of walking	81	11.11
D. Less than 30 minutes of jogging	0	0.00
E. 30–60 minutes of jogging	20	2.74
F. More than 60 minutes of jogging	4	0.55
G. Less than 30 minutes of off-leash play	42	5.76
H. 30–60 minutes of off-leash play	155	21.26
I. More than 60 minutes of off-leash play	135	18.52
J. No outdoor exercise	7	0.96
Q12. Do(es) your dog(s) perform any of the following physical activities regularly (at least once a week when the activity is in season)? Select all those that may apply. (n = 820)
A. Agility training	99	12.07	193.35	< 0.0001
B. Herding	15	1.83
C. Racing	11	1.34
D. Sled-pulling	13	1.59
E. Hunting/Retrieval	38	4.63
F. Other	145	17.68
G. None of the above	499	60.85
Q13. How active do you consider your dog(s)? If you have multiple dogs that vary in activity level, choose the level of activity of the most active dog. (n = 724)
A. Inactive	37	5.11	40.94	< 0.0001
B. Slightly active	152	20.99
C. Moderately active	360	49.72
D. Highly active	175	24.17

X² — Chi-squared.

Most respondents’ dogs did not participate in organized or judged events (499/820; P < 0.0001; Q12). For daily exercise, owners most frequently reported that dogs received 30 to 60 min of walking (195/729), followed by 30 to 60 min of off-leash play (155/729; P < 0.0001; Q11). Approximately 1% of respondents report that their dog(s) received no outdoor exercise on an average day (7/729; P < 0.0001; Q12). Respondents largely considered their dogs to be “moderately active” (360/724; P < 0.0001; Q13).

Feeding practices (Table 4)

Table 4.

Overview of owner reported diet provision and typical feeding practices of Canadian dog owners participating in an online survey about dog owner feeding practices.

Question	Frequency	Prevalence	X2	P-value
Q14. Do you feed your dog(s) a commercial dry dog food (kibble)? (n = 723)
A. Yes, solely kibble	282	39.00	28.53	< 0.0001
B. Yes, majority kibble	133	18.40
C. Yes, minority kibble	46	6.36
D. No	262	36.24
Q15. What do you feed your dogs in addition to kibble? (n = 179)
A. Commercial canned (wet) food	43	24.02	12.46	0.0142
B. Veterinarian-prescribed diet (dry or wet)	16	8.94
C. Commercial raw diet	31	17.32
D. Homemade diet	54	30.17
E. Other	35	19.55
Q16. Which of the following do you feed your dogs (if you do not feed kibble)? (n = 261)
A. Commercial canned (wet) food	1	0.38	60.26	< 0.0001
B. Veterinarian-prescribed diet (dry or wet)	56	21.46
C. Commercial raw diet	121	46.36
D. Homemade diet	59	22.61
E. Other	24	9.20
Q17. How frequently is/are your dog(s) fed? (n = 721)
A. Once daily	69	9.57	199.95	< 0.0001
B. Twice daily	550	76.28
C. Three times daily	45	6.24
D. More than three times daily	7	0.97
E. Food is always available	50	6.93
F. Other	0	0.00
Q22. When feeding your dog(s), do you: (n = 716)
A. Free-feed (no measuring)	76	10.61	140.11	< 0.0001
B. Measure the food given	636	88.83
C. Other	4	0.56
Q23. When measuring food, do you: (n = 633)
A. Measure using a non-scaled scoop or bowl	58	9.16	74.24	< 0.0001
B. Measure using a measuring cup	381	60.19
C. Weigh the food on a scale	157	24.8
D. Other	37	5.85
Q24. How do you decide how much food to offer your dog(s)? (n = 713)
A. Veterinarian recommendation	133	18.65	27.70	< 0.0001
B. Breeder recommendation	42	5.89
C. Feeding guidelines printed on the bag or can	188	26.37
D. Online food calculator	82	12.90
E. Other	258	36.19
Q28. Do you alter the amount of regular diet given to your dog(s) based on the body weight of your dog(s)? (n = 708)
A. Yes	573	80.93	103.54	< 0.0001
B. No	104	14.69
C. Unsure	31	4.38
Q29. Do(es) your dog(s) receive any commercial treats or table scraps in addition to their regular diet? (n = 708)
A. Never	46	6.50	27.55	< 0.0001
B. Rarely	128	18.08
C. Sometimes	275	38.84
D. Often	119	16.81
E. Daily	140	19.77
Q30. If on a given day you have provided treats or table scraps to your dog(s), do you then change the amount of the regular diet offered that day? (n = 654)
A. Yes	172	26.30	8.33	0.0155
B. No	307	46.94
C. Sometimes	175	26.76

X² — Chi-squared.

Over 63% of survey respondents reported feeding their dog(s) at least some kibble (461/723; P < 0.0001; Q14), and of those that fed another kind of food in addition to kibble, a homemade diet was the most common (54/179; P = 0.0142; Q15). For owners who reported feeding no kibble at all, over 46% fed a commercial raw diet instead (121/261; P < 0.0001; Q16). Owners most frequently responded “other” when asked how they decide the quantity of food to provide (258/713; P < 0.0001; Q24), with the second most frequent response being that they did use the feeding guidelines on the product packaging (188/713; P < 0.0001; Q24). Most respondents reported measuring the food provided (636/716; P < 0.0001; Q22) and of those, most use some form of measuring cup (381/633; P < 0.0001; Q23). Weighing food on a scale was the second most common measuring method (157/633; P < 0.0001; Q23). Most owners reported that they adjusted the amount of food provided according to their dog’s weight (573/708; P < 0.0001; Q28). Nearly 94% of respondents report feeding treats or table scraps at least occasionally (662/708; P < 0.0001; Q29), but almost half report not adjusting the amount of food offered in a day when these are provided (307/654; P = 0.0155; Q30).

Owner perception of body weight and body condition score (Table 5)

Table 5.

Overview of owner reported evaluation of the importance of weight maintenance and assessment of dog body condition score for Canadian dog owners participating in an online survey about dog owner feeding practices.

Question	Frequency	Prevalence	X2	P-value
Q31. How important is it to you that your dog(s) remain(s) at a particular recommended weight? (n = 705)
A. Not at all important	9	1.28	112.74	< 0.0001
B. Slightly important	42	5.96
C. Moderately important	171	24.26
D. Very important	483	68.51
Q32. If your dog(s) were to become overweight, would this concern you? (n = 705)
A. Not at all	2	0.28	149.16	< 0.0001
B. Slightly	29	4.11
C. Moderately	135	19.15
D. Very much	539	76.45
Q33. If your dog(s) were to become underweight, would this concern you? (n = 706)
A. Not at all	1	0.14	155.48	< 0.0001
B. Slightly	25	3.54
C. Moderately	132	18.70
D. Very much	548	77.62
Q34. If your dog(s) appeared to have gained weight, would you or those in your household do any of the following? Select all that may apply. (n = 1764)
A. Provide fewer treats or table scraps	489	27.72	39.24	< 0.0001
B. Provide less of the regular diet	405	22.96
C. Change diets	98	5.56
D. Seek veterinary advice	236	13.38
E. Change the exercise routine of the dog	497	28.17
F. Other	39	2.21
Q35. If your dog(s) appeared to have lost weight, would you or those in your household do any of the following? Select all that may apply. (n = 1278)
A. Provide more treats or table scraps	64	5.01	78.03	< 0.0001
B. Provide more of the regular diet	468	36.62
C. Change diets	144	11.27
D. Seek veterinary advice	483	37.79
E. Change the exercise routine of the dog	74	5.79
F. Other	45	3.52
Q36. Please select the image that best represents the body shape of your dog(s). Choose multiple if necessary, for multiple dogs. (n = 694)
A. BCS 3	317	36.90	64.99	< 0.0001
B. BCS 7	94	10.94
C. BCS 1	70	8.15
D. BCS 9	19	2.21
E. BCS 5	359	41.8

X² — Chi-squared. BCS — Body condition score.

Most respondents answered that it was “very important” that their dog(s) remain at a healthy body weight (483/705; P < 0.0001; Q31) and that they would be very concerned if their dog(s) became overweight (539/705; P < 0.0001; Q32) When asked what actions they might take if their dog(s) gained weight, owners reported they would provide fewer treats (489/1764), change the dog’s exercise routine (497/1764), or offer less of the regular diet (405/1764), whereas seeking veterinary advice was a less popular option (236/1764; P < 0.0001; Q34). When choosing a body condition score pictograph on a scale of 1 to 9 that reflected the body shape of their own dog(s), most respondents chose the image corresponding to a BCS of 5 (359/694), followed by the BCS 3 image (317/694; P < 0.0001; Q36).

Discussion

It was surprising that not all brands appear to use the modified Atwater method to calculate energy density, the only AAFCO approved method of calculating energy density, despite carrying an AAFCO statement on the label. The alternative is to determine the energy density through feeding trials. Both methods have inaccuracies, but the application of a standard method across the industry, either a calculation or experimental protocol, minimizes the ability to make superiority claims in addition to making it easier for owners to compare products and make informed decisions about their pet’s diet.

Due to the error generated when calculating energy density from the guaranteed analysis provided maximum and minimum values rather than analytical values from the proximate analysis, we assumed that larger percent differences between calculated and stated metabolizable energy are due to differences in the diet’s fat content, as fat is the largest contributor to the traditional and modified Atwater equations. Three products’ traditional Atwater calculation and 10 products’ modified Atwater calculation, calculated from the products’ guaranteed analysis, exactly matched the metabolizable energy content stated on the package (0% error). This is interesting because the average analytical results of the nutrient concentrations should be used for these calculations, not the values provided in the guaranteed analysis. Using the maximum and minimum values may introduce more error and give an inaccurate estimate of the metabolizable energy content compared to the analytical values.

Generally, feeding recommendations could be determined from the energy density on packaging and the estimated daily energy requirement of dogs. For most recommendations, it was clear which US NRC equations had been used in development. Percent errors at higher body weights were often larger compared to the lower body weights. The reason for this is unclear, but it could be that alternate equations and exponents are being used at higher body weights, as some US NRC equations may underestimate energy requirements (12).

Overall, it appears that the equations most frequently used to develop feeding guidelines are those that provide a low to moderate estimate of maintenance energy requirements for a given body weight. It was expected that performance products would use the highest estimating maintenance energy requirement equations, but this was not observed. Senior/weight management products infrequently used equations that provided a high estimate of energy requirements (E5, E6, E7), which is in line with products promoting a lower calorie intake. The equations used to develop feeding guidelines do not appear to be influenced by what breed size the product is marketed for. Together, these data suggest that dog food product labels are typically providing a conservative estimate of maintenance energy requirements that should be a reasonable starting point for food provision, as labels state. It is possible that the industry compensates for the modified Atwater method under-predicting energy density by using lower predictions for maintenance energy requirements. If this were the case, the feeding guidelines provided may indeed be very close to meeting the actual maintenance energy requirements of pet dogs. It has been noted that the process behind label feeding recommendations is inexact, resulting in a large variability (18). Any feed intake is a starting point and should be adjusted based on reassessment of body weight and body condition score and health outcomes (19).

The goal of this survey was to gain insight into the feeding practices of dog owners and their perceptions of their dogs’ body weight and body condition score. The responses show that owners’ reported concern for maintaining their dogs at an appropriate weight do align with the adjustments owners report that they would make to food, treat, and exercise provision if their dog(s) gained weight. Although respondents reported that they would be very concerned if their dog(s) became overweight, many would not prioritize seeking veterinary advice. This differs from previous findings indicating owners would seek veterinary advice first and are reluctant to decrease food or treat provision, which was veterinarians’ top advice (20,21). It also suggests that respondents understand the relationship between food, exercise, and their dog’s weight, and that they might feel confident addressing weight gain themselves before visiting a veterinarian.

Estimates are limited for a Canadian context, but the prevalence of dog obesity observed in this study, as based on owner report of body condition score, is much lower than what has been observed previously in other countries (22,23). Asking owners to choose the body condition score of their animal using a body condition score chart introduces several potential issues. Recall bias is possible if the dog is not present while the owner completes the survey, and inaccuracies may result from the dog’s shape differing from that in the pictographs. Owners are generally unreliable when estimating body condition score, even when using a body condition score chart (24). Underestimating or normalizing their dog’s body condition score may be indicative of an owner’s unwillingness to acknowledge obesity (24). One study showed that body condition score was accurately assessed by only half of dog owners, and 78% of those who incorrectly assessed had underestimated (18). Therefore, it is likely that at least some respondents underestimated their dog’s body condition score to result in the unexpectedly high proportion of dogs given a score of 3 or 5, especially considering dogs at a body condition score of 3 are in the underweight category. However, a body condition score of 3 is not necessarily a concern for sporting dogs, and it is possible that some respondents understand the value of being lean. It is also important to note that the circulation of this survey was done primarily via colleagues at the University of Guelph or social media groups that targeted owners of specific purebred dogs. Possibly, this method of distribution introduced a selection bias towards respondents who were more educated, and potentially more knowledgeable about and invested in their pet’s health. Owners in this study, however, placed a high importance on maintaining a healthy weight for their dogs, suggesting that they are largely appropriately managing their dog’s BCS in the home.

Nearly half of the respondents rated their dog(s) as moderately active, although we did not provide a description of this activity level out of interest in owners’ personal perception of dog activity. The high response rate for 30 to 60 min of off-leash play may encompass activities of widely varying intensity that owners interpret as off-leash play. Dogs with free yard access are prone to obesity (21), suggesting that owners inaccurately estimate or simply do not observe the actual amount of activity such dogs are voluntarily undertaking. A survey of Calgary dog owners found that while dogs living in attached housing, such as apartments, were walked more frequently than dogs with yard access, 43% of owners spent less than 150 min a week walking their dog, and very few owners exercise their dog at the level recommended for a given breed (25). Smaller breeds may be exercised less than medium or large breeds due to a lower perceived exercise requirement (26). In addition, product labels may provide feeding guidelines based on the activity level of the dog to be fed, for instance, “less active” and “more active,” a common observation in the feeding guidelines component of this study. When guidelines are presented in this way, owners may be led to overfeed due to an incorrect assessment of their dog’s true activity level.

Less than a third of respondents reported using feeding guidelines on a product package to determine how much to feed their dog. Of the options listed, veterinarian advice was the second most frequent way to determine how much food to give, unlike in other studies in which veterinarians were the primary source of nutritional information for owners (13,27). However, it is possible that veterinarians are not always emphasizing the importance of a healthy body weight out of discomfort addressing what is seen as a sensitive topic (28,29). A lack of body weight or body condition score discussion between client and veterinarian may prompt clients to seek information from less reputable sources or lead to a misunderstanding of the relationship between body weight and their dog’s overall health. The highest portion of respondents selected “other” when asked how they decide how much food to give, but it is unclear what these other methods are. It is possible this category includes such sources as the advice of other owners or pet retail staff, the Internet, adjusting food provision based on observation, or simply perceiving the dog as being hungry. Our survey, however, did not request this information. Future investigation should focus on the methods owners do use to determine how much to feed their dogs. Feeding a dog is often seen as a way to strengthen the human-animal bond, whereas denying food is viewed as a denial of love (30). Indeed, owner reluctance to enroll in weight loss programs for their dogs has been attributed to an unwillingness to withhold food (24), highlighting the importance of owner engagement in the success of any veterinarian-prescribed weight management plan.

Although it is promising that most owners reported measuring their dogs’ food, many used a measuring cup to do so. It is easy to overestimate the portion size by up to 80% using a measuring cup, and this is exacerbated for smaller dogs with smaller food and energy requirements (31). Even greater inaccuracy, up to 152% overestimation, has been recently reported (32). Scales that measure grams are more accurate and allow easy adjustments to portions. This is important for dogs maintaining their weight, and to ensure dogs which need to lose weight are doing so at a healthy rate. About 25% of respondents in this study claim to use a scale, corresponding to the high value these owners place on keeping their dogs at an ideal weight. If owners know the daily food intake that their dog requires, this is an excellent option. However, for some owners who rely on feeding guidelines, knowing this food intake in grams may be difficult if the guidelines are only presented in cups. Many product packages offer feeding guidance in both grams and cups, but this is not always the case, and may present a challenge for owners who might be willing to make the switch to a scale but are unsure about how.

Respondents indicated a preference for feeding twice daily, perhaps contributing to the higher than anticipated proportion of dogs identified at an ideal body condition score (21). However, most respondents reported providing treats or table scraps at least sometimes, without adjusting that day’s portion of regular diet to account for extra calories. This estimate aligns with previous reports that nearly all dog owners provide some form of treats some of the time (13,20). Providing treats without altering the regular diet portion is a risk factor for dog obesity (31,33). This sample of owners did, however, claim to alter regular food provision based on their dog’s body weight, suggesting awareness of the need to balance energy intake and expenditure for weight maintenance.

In summary, respondents to this survey tended to place high value on maintaining a healthy body weight and body condition score for their dogs and claimed that they would take appropriate actions such as reducing treats, increasing exercise, or reducing food provision to combat weight gain, indicating an awareness of the importance of healthy animal body weight. Canadian owners also appear to primarily be using sources other than commercial feeding guidelines and veterinary advice to determine how much food their dog requires. Investigation into what these specific methods are is important to determine how owners may successfully manage the BCS and overall health of their pets. Further studies should consider focusing on dog owners’ perceptions of pet food company feeding recommendations to determine the relationship these attitudes may have to pet health. Finally, we conclude that currently pet food companies are overall using the energy equations most practical for the dogs intended to be fed, but that there are instances of noncompliance with AAFCO regarding the method of calculating the energy density of the food. CVJ