Obesity in INTRODUCTION

the dog Obesity has been defined in man as an increase

in bodyweight of 10 to 1 5 per cent above the nor-
€? J. Markwell, W. van Erk*, G. D. Parkint, C. J.
mal weight for an individual (Craddock 1969).
Sloth#andT. Shantz-Christienson While this definition could be applied to dogs,
lack of baseline data for some purebred and all
crossbred animals, makes such application diffi-
Waltham Centre for Pet Nutrition, Waltham-on-the-Wolds,
Freeby Lane, Melton Mowbray, Leicestershire LE14 4RT, UK;
cult. An alternative approach which has been
*Vrouwgelenweg 26, Hendrik Ido, Ambacht, The Nether- applied to the study of obesity in man is the use
lands: tAshton House Veterinary Hospital, 69 Westgate End, of criteria such as skin fold thickness, girth
*
Wakefield, West Yorkshire, UK, and Skovlunde Dyreklinik, measurements and body density. However,
Skovlunde Byvej 53,2740 Skovlunde, Denmark when examined in relation to the dog, these
were found to be inappropriate or impractical
Journal ofSmal1 Animal Practice (1990) 31,533-537 (Anderson 1973).
As a result of this, most studies of obesity in
dogs have relied on assessment by largely subjec-
ABSTRACT ~~
tive criteria. In spite of this limitation a number
of interesting surveys have been conducted in
Canine obesity is a common condition in vet- practice, and the results of these are given in
erinary practice, and has been associated with a Table 1. Because these surveys were conducted
number of serious medical conditions likely to in veterinary practice the numbers may not be
reduce the length and quality of a pet’s life. representative of the population as a whole, but
Therefore, it represents an important condition it seems reasonable to conclude that between
for the veterinarian to manage. This paper reports one-quarter and one-third of dogs presented are
a study conducted by veterinarians in practice overweight.
using controlled calorie reduction for the man- Certain subpopulations of dogs appear particu-
agement of overweight dogs. In the study weight larly likely to be overweight. Edney and Smith
losses of 14 per cent were achieved over a 1 2 (1986), in their extensive survey of dogs pre-
week period. Details of this study are discussed. sented to practices, found that neutered females
were about twice as likely to be obese as entire
females, and a similar trend was evident in
neutered males. Certain breeds also appeared
particularly prone to obesity, including the
labrador retriever, Cairn, cocker spaniel, long-
haired dachshund, Shetland sheepdog, cavalier
King Charles spaniel and beagle.
An interesting, and alternative definition of
obesity, is that it is a pathological condition char-
acterised by an accumulation of fat much in
excess of that necessary for optimal body func-
tion (Mayer 1973). Implicit in this definition is
the concept of obesity as a condition detrimental
to an individual’s health. The lack of objective
studies in dogs makes a clear demonstration
of a link between obesity and disease dif-
ficult, and many more conditions have been
implicated as linked with obesity than have been
Table 1. Obesity in dogs

Percentage of dogs, presented to veterinary practice,

classified as obese
I I
Mason (1970)
Anderson (1973)
Steininger (1981)
Edney and Smith (7986)
This paper was presented at the Waltham symposium number
*Includes dogs categorised as obese or gross
13,1989

533
 P. J. MARKWELL AND OTHERS

demonstrated. However, canine obesity appears was determined on an individual basis. Results
to be linked with a decreased resistance to infec- from a series of dogs targeted for bodyweight loss
tious disease (Newberne 1966), diabetes mellitus, of up to 16 per cent of starting weight will be
articulo-locomotor problems (Edney 1974), circu- considered here.
latory problems, at least in dogs classified as The dogs were prescribed a low calorie, com-
gross [Edney and Smith 1986), and increased sur- plete canned food (Canine Low Calorie Diet;
gical risk. Other conditions may well be linked Effem, Austria]. The amount of this diet allocated
with obesity but Edney and Smith (1986) failed represented approximately 40 per cent of the cal-
to demonstrate a relationship between obesity culated maintenance requirement of the dog at its
and skin or reproductive problems or neoplasia. target weight. It was considered impractical to
On the basis of these studies it seems reasonable expect owners to accurately weigh food and
to suggest that obesity is detrimental to the therefore the amounts to be offered to the dog
health of the dog and may result in shortening of were approximated to the smallest unit of one-
the dog’s life and reduction in its quality of life. quarter of a 410 g can. The exception to this was
Obesity is actually caused by the dog’s energy for two very small dogs, where one-third and
intake exceeding its requirement at some stage two-fifths of a can were recommended. Several
during its life. Positive energy balance results in dogs were also offered lower amounts, but the
the dynamic phase of obesity, which may be fol- results for these will not be considered further in
lowed by the static phase, where the dog is over- this report.
weight but no longer gaining weight. Obesity has Owners were given a letter reminding them of
been associated with a wide range of endocrine the importance of obesity and of following the
and other problems in man, although these prob- regimen accurately. They were also asked to com-
ably account for a minority of cases. Armstrong plete a simple record chart detailing their dogs
and others (1951) considered that the excess daily food intake, on the basis of 0 , one-quarter,
weight, in at least 95 per cent of cases, was one-half, three-quarters or all of the food offered
accounted for by overeating. While similar data being eaten. They were asked to return to the
do not exist for dogs it is unlikely that the pro- clinic every two weeks for the dog to be
portion of cases associated with endocrine causes reweighed, and for discussion and assessment of
is greater. Therefore, the vast majority of cases of the dog’s progress and the level of feeding.
canine obesity encountered by the veterinarian Where it was not possible for the owners to
are likely to result from overfeeding. return at precisely two week intervals the
The frequency with which canine obesity is weights recorded have been approximated to the
observed, and its association with serious medi- nearest two week period.
cal conditions, ensure that it is an important Assessments of the rates of weight loss of
challenge to the veterinarian. The fact that most neutered and entire bitches and dogs were made
cases are associated with poor feeding practices using regression analysis of individual body-
means that a programme of controlled calorie weight changes. Regression coefficients were
reduction can be utilised for management of compared using Student’s t test.
overweight dogs. The objective of this paper is to
report the results obtained with such a
programme. RESULTS
Results from 34 dogs targeted for a bodyweight
CLINICAL STUDY loss of 16 per cent or less, and followed for up to
1 2 weeks, are presented in Table 2. This table also
This study was conducted by veterinarians in gives details of the breed, age, sex and target
three private practices and was designed to weights for each of the dogs. The starting weights
reflect what could be achieved in practice, rather varied from 3.5 to 48.4 kg and the targets for
than under laboratory conditions. The protocol weight loss from 7.3 to 15.9 per cent of initial
used was similar to that described by Edney bodyweight. The ages of the dogs varied from one
(1974) which had been used for the successful year to over 11 years. During the period of the
management of obesity in a series of dogs. Each study 20 of the dogs reached or passed their target
dog was recruited for the trial based on the clini- weights, although on one occasion a small weight
cians clinical examination and on their subjec- gain was noted on a subsequent weighing. The
tive assessment of the dog’s body condition. The table also shows the mean and standard deviation
owner was then counselled about the problems of the bodyweight changes of the dogs being
associated with canine obesity and their coopera- weighed on each occasion during the study
tion sought for the trial. The importance of obser- period. The results show a mean loss of 14 per
vance of the protocol was stressed to the owner. cent of initial weight had been achieved by the 24
The dog was then weighed and a target weight dogs still being followed at the end of the study.

534
 Obesity in the dog

Two of the dogs which had not achieved their whether the animal was entire or neutered. Statis-
target weights were followed for an extended tical comparison of the data for the entire bitches
period. One of these dogs, for which no record of and the spayed bitches showed that there was no
food intake is available, achieved its target significant difference between these two groups
weight loss (14.1per cent) by the 22nd week. The in their rate of weight loss. A similar comparison
other dog was followed for 15 weeks and by this of the entire and castrated dogs again showed that
time it had only lost 9-2 per cent of its starting there was no significant difference between the
weight. The owner’s record indicated that the rates of weight loss of these two groups.
dog had frequently eaten less than its full
allowance of food, so a greater rate of weight loss
might have been expected to occur. However, the DISCUSSION
owner also recorded food being stolen by the dog
on two occasions, so it is possible that the low The results from this study show that substan-
rate of observed weight loss could have been due tial weight reduction can be achieved in over-
to more frequent (but unobserved) theft of food. weight dogs, in a practice situation, using a
Table 3 presents data from the 31 dogs for programme of controlled calorie reduction. The
which information is available, detailing sex and results also indicate that it is reasonable to

Table 2. Data from dogs in weight reduction trial

I Age of Percentage of starting weight after

i
Enti dog Start 2 4 6 8 10 12 Target
Breed Sex neut yearsrnonths wt(kg) weeks weeks weeks weeks weeks weeks weight (kg)
~~

Yorkshire terrier M E 9 3.5 91.4 91.4 82.9 3.0t

Pomeranian F E 7 3.95 89.6 81.8 80.5 75.9 75.7 72.2 3.35
Yorkshire terrier F N 7 7 5.1 96.1 94.1 92.2 86.3 86.3 4.4
Poodle F N 10 6.4 90.6 89.1 84.4 82.8 5.4
Terrier F E 9 10.6 94.3 90.6 86.8 83.0 83.0 80.2 9.0
West Highland white F N 9 10.8 95.4 93.5 91.7 91.7 88.9 88.0 9.1
Pomeranian M E 10 11.4 96.5 93.0 89.5 91.2 87.7 87.7 9.7*
Poodle F N 12 11.6 96.6 93.1 93.1 89.7 87.9 89.7 9.9*
Schnauzer M E 2 12.4 91.9 88,7 83.9 82.3 79.0 77.4 10.6
Crossbreed F N 8 6 16.7 95.8 92.8 91.6 91.6 88.0 15.0
Cocker spaniel M E 5 7 17.0 96.5 91.2 84.7 84.1 82.4 14.4
Spaniel F N NS 17.1 92.4 91.2 86.5 83.0 81.3 79.5 14.5
Americancocker F E 5 5 18-4 95.7 94.6 92.4 93.5 93.5 90.2 15.8
Border collie F N 6 19.1 96.3 96.3 90.1 86.4 17.3
Fox terrier M C 8 21.2 95.3 91.0 92.5 91.0 87.7 88.7 18.0+
Crossbreed F N 7 21.3 96.7 93.9 93.0 91.1 89.2 19.1
Rough collie M E 2 6 28.3 95.8 93.6 91.2 87.6 25.0
Collie F NS 9 28.6 98.6 92.3 88.1 83.2 24.3
Cocker spaniel M C 2 29.4 89.8 86.4 87.1 81.6 78.2 78.2 25.0
Retriever F E 7 30.8 94.2 94.5 90.9 88.3 88.0 84.4 27.0
Collie F E 6 31.2 98.7 94.9 92.9 91.7 90.4 89.1 26.5
Labrador F N 4 31.2 95.5 92.9 90.4 26.5
Labrador F N 11 6 31.3 99.0 97.4 94.9 94.9 93.9 90.4 28.2
Labrador M C 11 11 31.4 97.8 95.5 94.3 92.4 29.1
La brador F NS 1 32.8 99.1 96.6 94.8 94.5 92.1 91.2 28.2
Labrador cross M C 1 4 35.0 96.9 92.0 90.6 31.8
Dobermann F N 4 36.2 94.5 92.3 89.5 89.0 84.5 86.7 31.0
Springerspaniel M C NS 36.5 97.5 95.3 92.6 91.5 89.9 86.6 31.0
Springerspaniel M C 6 2 38.9 95.1 95.4 95.3 95.1 92.0 90.7 32.7
Collie M C 5 40.2 95.0 94.0 90.0 87.6 85.1 84.6 34.0
Retriever F N 11 41.6 98.1 96.9 93.8 92.8 90.9 90.4 35.0*
Retriever M NS 4 6 45.4 97.4 97.4 96.5 94.7 94.7 95.2 39,0*
Boxer/la brador M C 4 46.7 92.5 89.1 87.8 85.2 84.4 81.4 39.7
Retriever M E 9 48.4 98.1 96.9 95.9 92.1 92.6 91.7 41.0

n= 34.0 34.0 34.0 32.0 30.0 27.0 24.0 34.0

Mean = 25.0 95.4 92.9 90.4 89.1 87.0 86.0 21.6
SD= 12.9 2.6 3.3 4.0 4.6 4.9 5.4 11.1
I

*Food allowance reduced during study M=Male C=Castrated

tWeighing stopped after six weeks by error F=Female N=Neutered
E=Entire NS=Not stated

535
 P. J , MARKWELL AND OTHERS

Table 3. Percentage weight changes in dogs during weight reduction trial - results grouped by sex
Percentage of starting weight (fSD)
I
Weeks
on trial
Entire
females
Neutered
females
Entire
males
Neutered
males I
I 0 100 (5) 100 (12) 100 (6) 100 (8)
i
2 +
94.5 2.9 (5) 95.6f 2.2 (12) 95.0 5 2.5 (6) 95.0 k 2.5 (8)
4 91.3 f 5.0 (5) 93.6? 2.3 (12) 92.5f 2.5 (6) 92.35 3.1 (8)
6 88.7+- 4.6 (5) 90.85 3.0 (11) 88.0 f 4.6 (6) 91.3f2.7 (8)
8 86.5f 6.4 (5) 90.25 3.3 (10) 87.5 k 3.8 (5) 89.2f4.3 (7)
10 86.1 f 6.2 (5) 87.4& 3.6 (10) 85.4k 5.2 (4) 86.2f 4.4 (6)
12 83.2? 6-6 (5) 87.3 3.5 (7) 85.6f 6.0 (3) 85.0k 4.2 (6)

i L u m b e r s i n brackets refer t o number of dogs weighed at given times

I

expect most overweight dogs to achieve on aver- the range considered here and fed as described,
age a weight loss of approximately 14 per cent started the trial but stopped after varying periods
over a 12 week period when following this regi- of time. Three of these dogs had actually lost
men. This information is clearly of value to the weight initially before weight gain was noted,
veterinarian at the beginning of the trial when presumably because the owners started but then
discussing the condition with the owner, and abandoned the feeding regimen. One additional
later, when monitoring the progress of the dog. dog selected for inclusion in this trial refused to
A considerable degree of individual variation eat the food over the first four days and was
in weight loss was expected, and has been noted withdrawn. One other dog was actually fed more
in previous trials (A. T. B. Edney, personal com- by the owner than the food allowance in the lat-
munication). Much of this variation is likely to ter part of the trial, although it still achieved
be associated with differences in energy require- its target weight loss. Overall this, and other
ment between individuals, due to factors such as trials (A. T. B. Edney, personal communication),
temperament and amount of exercise. Some of indicate a number of factors that may help in
the variation was likely to have been associated contributing to owner compliance.
with differences in food intake, and the owners’ These factors include a detailed initial consul-
records, although providing very approximate tation, discussing the clinical problems associ-
data, can be of considerable value to the veteri- ated with obesity and the importance of its
narian in this respect. Another factor to be con- control. A realistic initial target, and guidance as
sidered, is that in a practice situation it is not to the time expected to be taken in reaching that
always possible to weigh a dog at the same time target is also important. Too large an initial target
of day on each occasion, although it is desirable may lead to the owner giving up, whereas once
to do so. These factors may account not only for one target is achieved and the benefits noted,
some of the variation seen, but also for the level- another can always be set. The regular revisits to
ling off of weight loss (or slight weight gain) the surgery are also of value, both to monitor
noted at various times during the trial in approxi- progress and to reinforce the importance of
mately half of the dogs, and it may explain why, adherence to the feeding regimen. Use of a sim-
in many cases, further weight losses were noted ple graph to follow weight changes may be bene-
without reduction in the food allowance. ficial, particularly in the early stages, where
The owners’ records can provide valuable data changes in the dog may not be very evident. Use
about food intake when reassessing cases and of a palatable diet is also valuable, as the owner
also when considering reductions in food is likely to abandon the regimen if the dog
allowance. In some of the cases noted here, refuses the food.
reduction of food offered to the dog actually The comparison of weight losses of neutered
made little difference to its overall recorded and entire dogs and bitches were interesting, par-
intake. Indeed in one case an increase was noted ticularly considering the much greater propensity
from an average consumption of approximately for neutered animals to be obese (Edney and
0.7 cans per day to one can per day when the Smith 1986). The reasons for this increased risk
amount of food offered to the dog was reduced! of obesity are not entirely clear, but the factors
This also serves to emphasise the importance of that are involved might have been expected to
individual assessment of cases. contribute to a slower rate of weight loss in the
Obtaining and maintaining the owners’ interest neutered animals. Although this was not
and enthusiasm for the trial is a key factor in suc- observed in this study, further investigation in
cess. Owner compliance was generally good, the area of energy requirement and expenditure
although an additional four dogs targeted within of neutered dogs is clearly needed.

536
 Obesity in the dog

nal of the American Medical Association 147, 1007-1014

CONCLUSION CRADDOCK D. (1969)The size of the problem. In Obesity and

its Management. Livingstone, Edinburgh. p 1
EDNEY, A. T. B. (1974)Management of obesity in the dog. Vet-
erinary Medicine/Small Animal Clinician 49, 46-49
In summary, canine obesity is an important EDNEY, A. T. B. & SMITH,P. M. (1986)Study of obesity in dogs
clinical challenge to the veterinarian. This study visiting veterinary practices in the United Kingdom. Veteri-
has shown that it can be managed successfully in nary Record 118,391-396
a practice environment by a programme of con- MASON,E. (1970)Obesity in pet dogs. Veterinary Record 86,
trolled calorie reduction. 612-616
MAYER, J. (1973)Obesity, In Modern Nutrition in Health and
Disease. Eds R.S. Goodhart and M.E. Shils. Lea and
Febiger, Philadelphia. p 625
REFERENCES ~
NEWBERNE, P. M. (1966)Overnutrition and resistance of dogs
to distemper virus. Proceedings of the Federation of the
American Society of Experimental Biology [FederationPro-
ANDERSON, R. S. (1973)Obesity in the dog and cat. Veterinary ceedings) 25, 1701-1710
Annual, 14th edn. J. Wright, Bristol. p 182 STELNINGER, E. (1981) Die Adipositas und ihre diatetische
ARMSTRONG, D. B.,DUNLIN, L. I., W m m , G. H. & MARKS, H. H. Behandlung.. Wiener Tierarztlicher Monatschrift 88, 122-
(1951) Obesity and its relation to health and disease. Jour- 130

ABSTRACTS common in women. ATA values can help in the

diagnosis of hypothyroidism in dogs.
BEALE,K. M., HALLIWELL,R. E. W. & CHEN,C.L. (1990)Journal
Prevalence of antithyroglobulin of the American VeterinaryMedical Association 196, 745
antibodies detected by ELISA of
canine serum Management of tracheal stenosis
HYPOTHYROIDISM is the most common in a dog
endocrinopathy in dogs. One cause of hypothy-
roidism is lymphocytic thyroiditis, as a result of A TWENTY-eight-month-old cocker spaniel pres-
which antithyroglobulin antibody (ATA) is pro- ented with breathing difficulties secondary to
duced. This study investigated the prevalence of tracheal laceration which had been repaired sur-
ATA in a large population of dogs suspected of gically. One month after the repair the dog had
hypothyroidism, and in two groups of healthy progressive exercise intolerance, dyspnoea and a
dogs. The ATA values were higher in dogs with moist, productive cough. Radiographs demon-
low concentrations of total serum thyroxine (T,) strated narrowing of the mid-cervical trachea.
and triiodothyronine (T,) than in dogs with Three bougienage procedures performed at 10
normal values of T, and T,. In the population day intervals failed to correct the stenosis. Resec-
studied dobermann pinschers, which are known tion of the stenosis with tracheal anastomosis
to be predisposed to hypothyroidism, had an improved the diameter of the trachea and allevi-
abnormally high incidence of ATA (20 per cent). ated the clinical signs.
There was no association between the sex of the SMITH,M. M., GOURLEY, J. M., AMIS,T. C. & KERPERSCHOEKEG,
C.
dogs and the incidence of ATA, whereas in (1990)Journal of the American VeterinaryMedical Associa-
human beings lymphocytic thyroiditis is more tion 196,931

537