Canine Obesity

Canine obesity is a condition that leads to serious alterations in various bodily functions and limits the longevity of the animal. It is the most common nutrition-related canine disease in industrial countries, but its treatment is effective. The prevention of obesity demands nutritional discipline throughout the life of the animal and specifically at certain key times when the risk of becoming overweight is increased. The treatment of canine obesity is complicated by various problems, including a lack of awareness among owners, who often fail to recognize or minimize their animal's obesity. Without the total cooperation of owners, it will be impossible to obtain weight loss in the dog. The veterinarian will therefore first have to convince the owner before implementing weight loss protocols and obese animal care.

While various stages or degrees of weight gain can be distinguished in dogs, in this summary we use the term "obesity" to qualify any pathological weight gain. We strive to define obesity, to explain its pathogenesis and related problems, its evaluation, the various types of diet and the practical treatment of this disease.

Marianne DIEZ
DVM, Ph.D, Dipl ECVCN

Marianne Diez is a doctor of veterinary medicine and an associate professor in the Domesticated Animal Nutrition Department of the Veterinary Medicine Faculty of the Université of Liège (Belgium), where she obtained her PhD in 1998. She is responsible for general nutrition and companion animal nutrition. She is committed to the development of nutritional consultations for pet owners. Her main area of focus is canine nutrition, specifically dietary fiber, prebiotics and canine obesity. She is the author or co-author of approximately 80 scientific articles and a Founding Diplomate of the European College of Comparative Veterinary Nutrition (ECVCN).

Patrick NGUYEN
DVM, MS, Dipl ECVCN

After graduating from the Ecole Nationale Vétérinaire of Alfort in 1977, Patrick Nguyen worked as an Assistant in the Department of Nutrition for two years before joining the Ecole Nationale Vétérinaire of Nantes. In 1982 he passed his Agrégation in Nutrition to become a teacher. He obtained an Authorization to Direct Research at the Université de Nantes and became a Diplomate of the European College of Comparative Veterinary Nutrition (ECVCN). Principle assessor of the Ecole Vétérinaire of Nantes in 1992 - 1996, he became the head of the Nutrition and Food Unit, and has also headed the Biology and Pharmacology Department since 2001. Patrick is responsible for the teaching and research programs in nutrition. His primary field of research is obesity and insulin resistance in companion animals, in association with the Centre de Recherche de Nutrition Humaine in Nantes. He is also involved in research projects on the effects of neutering in cats and on the digestive sensitivity of large-breed dogs. Patrick has more than 100 publications and papers. He became the President of ECVCN in 2004.

1. Canine Obesity

Definition

Obesity is a "pathological condition characterized by excessive fat deposition leading to modifications to various bodily functions". The World Health Organization (WHO, 1997) goes even further, defining human obesity as "excessive fat leading to harmful consequences for health". Although this definition appears fairly brutal, it can certainly be extrapolated to companion animals. At a quantitative level, obesity is described as being 15% overweight as compared to optimal weight. This somewhat reduced approach is no longer used as such; it has been replaced by body mass indices, which determine an optimal weight range for men and women of a given size. No such tool exists for dogs. A "mathematical" definition of obesity is of little use (Markwell & Butterwick, 1994) as it requires knowledge of the healthy weight, which, even for pure-breed dogs is not always easy to determine. The ideal situation is to know the weight of the adult dog before it became obese. This body weight is used as a benchmark in both the initial and follow-up evaluations of the animal. In some cases, the optimal body weight is unknown as the animal has always been overweight, even during the growth phase.

Body condition scores are practical, specific tools that have been developed for veterinarians (see below). A dog is obese when its body condition score is more than 3 on a 5-grade scale. (© Alex German).

Epidemiology of Obesity

Frequency

In the most recent studies, the frequency of obesity in dogs presented at consultation varies from 24% to 44% depending on the author (Table 1), the epidemiological study location and the predefined criteria (Mason, 1970; Meyer et al., 1978; Edney & Smith, 1986; Armstrong & Lund, 1996; Robertson, 2003).

These data do not always reflect the local situation. Some studies continue to serve as a reference despite being over 30 years old, while others have been conducted in a limited number of veterinary clinics and do not necessarily reflect differences between countries. Nevertheless, all studies conducted in veterinary consulting rooms in industrialized countries and large cities show a prevalence of obese dogs of at least 20%. A telephone survey conducted among 400 veterinarians in four European countries (France, Germany, Spain and the United Kingdom) in May 2000 showed that these veterinarians estimate the proportion of obese dogs (where obese is associated with the necessity of implementing a low calorie diet) to be 20% (Royal Canin, 2000).

In conclusion, the epidemiological data does not prove that the frequency of obesity has been increasing for 10 years, but obesity is still a major medical problem in the canine population.

Risk Factors (Table 2)

Breed

Breed is an obesity risk factor in the canine species but the predisposed breeds vary depending on the author and the study. For example, in the United Kingdom the Labrador Retriever, the Cairn Terrier, the Collie, the Basset Hound, the Cavalier King Charles Spaniel, the Cocker Spaniel, the longhaired Dachshund and the Beagle were predisposed and often cited in the 1980s (Edney & Smith, 1986).

These breeds were very popular in the United Kingdom when the study was conducted. The breeds that suffer may vary depending on the country and certain other factors. Krook et al. (1960) report that the breeds predisposed to obesity in Sweden were the Rottweiler, the Saint Bernard, the Collie, the Newfoundland, the Scottish terrier and the Chow Chow. Conversely, some breeds - sight-hounds and sheepdogs - appear to be more resistant. In a study conducted by Meyer et al. (1978) in Germany, German Shepherds, Poodles and Boxers were frequently obese. It is therefore necessary to qualify this idea of breed predisposition, even though practitioners do state that they encounter more obese Labrador Retrievers than obese sight-hounds. Obesity also appears to be a problem in other breeds (Table 3). The type of selection may influence the physical condition (and the weight) of dogs, for example, by substituting beauty or size criteria for working aptitude. Breed predispositions are partly related to genetic factors, more specifically in the fat mass/lean mass ratio that determines the maintenance energy requirement (Figure 1).

Figure 1. Comparison of the distribution of lean and fat mass in several large-breed dogs (Royal Canin, 2003-2004).

Not all breeds are the same when it comes to nutritional risk during growth. The energy excess predisposes small-breed dogs to obesity, whereas the major risk among large breeds is osteoarticular complaints (Grandjean & Paragon, 1996). The association of articular problems and obesity is frequent at the end of the growth phase of large breeds.

The Bernese Mountain Dog is one of the giant breeds predisposed to developing obesity. (© Renner).

The Collie is one of the medium breeds most at risk from obesity.
The Labrador Retriever is one of the large breeds most at risk from obesity. (© Hermeline).

The Cairn Terrier and the King Charles Spaniel are among the small dogs predisposed. (© Renner).

Genetic Factors

The purpose of a complex system of genetically determined factors is to maintain balance between dietary intake and energy expenditure. These regulatory mechanisms are particularly well adapted to helping wild species survive in times of famine. Nevertheless, when food is abundant, these factors do not appear to maintain a balance between intake and expenditure in domesticated animals in a confined environment, which has led to an increase in the obese population. Whatever the case may be, some individuals become obese while others living in the same conditions maintain their ideal weight. So it is not easy to distinguish between environmental factors in a wide sense and genetic predisposition (Johnson, 2000).

Genetic factors leading to obesity in dogs are still poorly understood. It is however, undeniable that these factors play a role since obesity is particularly common in selected breeds of dogs and in certain lines. The polygenic nature of obesity is undisputed (Schalling et al., 1999).

Age

The frequency of obesity increases with the age of the dog (Robertson, 2003) and the age of the owner (Edney & Smith, 1986). It is found in only 6% of puppies age 9-12 months, but in 40% of adult individuals (Glickman et al., 1995). The average age of diagnosis varies from 5 to 8 years. Less than 20% of dogs age 4 years and younger are obese, although that rises to over 50% in the 7-8-year category and to almost 70% for the 9-years-and-older category (Meyer et al., 1978) (Table 4). In very old dogs there is a contradiction between the figures cited by Mason (1970) and more recent data that shows that the frequency of obesity declines in old dogs age over 12 years (Armstrong & Lund, 1996).

An overweight puppy is predisposed to obesity in adulthood. (© Doxicat).

Obese bitches between 9 and 12 months are 1.5 times more likely to become obese adults than lean bitches during growth (Glickman et al., 1995). By way of comparison, in humans, 80% of obese adolescents become obese adults and they are also fatter than those adults that become obese without having being overweight during growth (Abraham & Nordseick, 1960). These data are confirmed by epidemiological studies (Eriksson et al., 2003).

Sex

The various data presented in Table 5 show that females are more predisposed to obesity than males. In certain studies, females represent more than 60% of obese dogs (Krook et al., 1960; Jerico & Scheffer, 2002). Furthermore, Glickman et al. (1995) have observed an obesity percentage of 40% in a study of 289 adult dogs.

Neutering

Gonadectomy increases the frequency of obesity in males and especially female dogs (Anderson, 1973; Edney, 1974; Karczewski et al., 1987; Miyake et al., 1988; Robertson, 2003). Edney and Smith (1986) have observed that neutered bitches were twice as likely to be obese compared with intact bitches. A more recent study shows that this is also true for male dogs. The frequency of obesity is 32% and 15% in neutered and unneutered animals, respectively, males and females together (Robertson, 2003). The sexual hormones are not prime regulators of metabolism, but they nevertheless have a direct impact on body weight at the level of the central nervous system, or indirectly by modifying cellular metabolism. Moreover, estrogens exercise an inhibitor effect on feeding. Dietary consumption thus varies in the female's sexual cycle. It is minimal during estrus, increases during metestrus and is maximal during anestrus (Houpt et al., 1979).

Various studies show that females are more predisposed to obesity than males. (© Alex German).

The influence of premature neutering on the incidence of obesity is not well known. An American epidemiological study reports that the frequency of obesity is lower in a population of dogs neutered before 5.5 months than in animals neutered between 5.5 and 12 months. The authors also report an overall incidence of obesity of 27% in the neutered population (Spain et al., 2004).

While it is difficult to clarify the link between neutering and obesity due to its multifactorial nature, several explanations may be considered. The first point to consider is the variation of food ingestion during the sexual cycle as stated above and the inhibitor effect of estrogens on food consumption. It is logical to suppose that this inhibitor effect is no longer exercised in neutered females. For a period of three months following neutering, four female Beagles consumed 20% more food than the unneutered control animals and their weight increased significantly (Houpt et al., 1979). Another study addressed this problem by measuring not only the weight gain in the neutered females, but also the quantities of energy needed to maintain a body weight considered ideal for Beagle bitches. A 30% reduction in daily energy intake compared with intake before neutering has been shown to be necessary to maintain bitches at their ideal weight in the weeks following an ovario-hysterectomy (Jeusette et al., 2004a). This level of energy restriction appears to be high, but one of the explanations advanced is that the Beagle is particularly predisposed to obesity. Neutering also leads to a reduction in spontaneous activity, especially among males. This last point is difficult to quantify in the breeding kennel.

The weight gain after neutering may accordingly be prevented with strict dietary measures and regular exercise. In a study on German Shepherds trained on an obstacle course and used as patrol dogs, no difference in body weight was apparent between neutered and unneutered bitches, although they all received the same quantity of food (Le Roux, 1983). This fact demonstrates that the maintenance of regular exercise after neutering can prevent weight gain.

The scale of neutering in the canine population may explain the increase in the frequency of obesity since the first epidemiological studies were reported in 1960. Moreover, as the practice of neutering becomes more and more widespread, we must expect an increase in the frequency of obesity in the coming years, particularly in those countries that have yet to be affected by obesity.

Contraceptive Treatments

Contraceptive treatment with medroxyprogesterone acetate led to a significant weight gain in 17.4% of the bitches treated during a clinical trial. The authors reported bulimia and obesity in some animals (Picavet & Le Bobinnec, 1994). The increase in weight following this contraceptive treatment is well documented in the bitch (Harel et al., 1996).

Obesity and Endocrine Diseases

Obesity may be associated with certain endocrine diseases, such as diabetes mellitus (Krook et al., 1960; Mattheeuws et al., 1984a; Wolfscheimer, 1990; Ford et al., 1993; Hoenig, 2002) and hypothyroidism (Kaelin et al., 1986; Forbes & White, 1987; Roche et al., 1991; Ford et al., 1993; Panciera, 1994, 2001; Dixon et al., 1999). According to the authors, at least 40% of bitches suffering from these conditions are obese. Obesity may also be secondary to hyperadrenocorticism. In a clinical study, five dogs presented fat deposits typical of obesity and different from a pendulous abdomen (Spearman & Little, 1978).

Obesity Secondary to Medication

Some medications may lead to hyperphagia and weight gain, particularly antiepileptics and glucocorticoids.

Sedentary Lifestyle and Lack of Exercise

Lack of exercise is a primary factor in the development of obesity: the prevalence of obesity declines in proportion to the duration of daily exercise. It is not possible to establish whether obesity is responsible for restricting exercise or whether the lack of exercise constitutes one of the factors responsible for obesity (Robertson, 2003). The duration of daily exercise is a more precise criteria than the type of habitat when evaluating energy expenditure.

Generally speaking, there are more obese animals among dogs living in an apartment than among dogs living outdoors (31% versus 23%) (Robertson, 2003). Nevertheless, it would be a mistake to believe that access to an open play space systematically increases energy expenditure. Some animals that live in a confined environment walk for several hours a week, while others that have access to a garden are content with a few minutes a day.

German Shepherd. Regular physical exercise is an effective way of preventing obesity. (© Renner).

Type of Food

The following dietary habits have been clearly identified to contribute to obesity: dietary intake that does not take account of the energy requirements ("the dog eats everything it's given") and supplements in the form of treats or snacks not accounted for in the energy intake. Highly palatable food high in easily assimilated fats and carbohydrates also predispose a dog to obesity. An undeniable risk factor is ad libitum feeding leading to overconsumption of energy.

Food can be highly palatable due to the presence of aromas or large quantities of fat. Foods with the highest fat content are also those with the highest energy concentration. Although it tolerates and utilizes fats as sources of energy, the dog is also able to immediately store them in the form of abdominal fats.

Altering the nutrient composition of a diet by an 8% increase in fat, without modifying the total energy intake resulted in a significant increase in the deposition of abdominal fat without modifying body weight in a group of bitches (Kim et al., 2003). In humans the fat intake is the main determination in the development of obesity (Garaulet et al., 2001). In dogs, highly digestible food - low in dietary fiber and with a very high energy concentration - may also be responsible for weight gain. Various treats, leftovers and nutritional supplements are additional risk factors (Kienzle et al., 1998; Robertson, 2003).

Dachshund. Many small dogs spend the majority of their time in the home. In Asia, 65% of small dogs have a 100% indoor lifestyle: these dogs do not go outside and use a litter tray in the same way as a cat does. (© Doxicat/Hermeline).

There is disagreement about the influence of home-prepared diets on the development of canine obesity (Lewis, 1978). The underlying idea is that the dogs that receive home-prepared rations will most often be compensated with treats and will receive larger quantities of food. This can only be defended in countries where dogs continue to be fed in the traditional way, with home-prepared rations or leftovers. In North America, 95% of animals are given commercial food and yet canine obesity is at least as widespread as in some European countries (Lund et al., 1999).

An epidemiological study has not shown the particular influence of a type of food (wet versus dry) on the frequency of obesity (Robertson, 2003).

Female Auvergne Pointing Dog. Hypothyroidism is often associated with obesity. (© Diez).

Contrary to some preconceived ideas, dividing the daily diet into several meals does not lead to an increase in the frequency of obesity. In epidemiological studies obese dogs are generally fed once a day (Kienzle et al., 1998; Robertson, 2003). It is clearly important not to confuse the division of the appropriate daily ration with the multiplication of the additional treats.

Social Aspect of the Food

The place of food in the relationship between the human and the dog plays a major role in the development of obesity.

Among the sociological factors, a study conducted in Germany (Kienzle et al., 1998) reported that the relationship between the human and the obese dog is characterized by excessive anthropomorphical behavior. For example, owners of obese animals talk to their dog more, allow their dog to lie on their bed, are not concerned about zoonoses and consider exercise, work and the protector function of their dog to be of minor importance. So it is unsurprising that obese animals are given meals or treats more often than animals of normal body weight. This study confirms that the owners are often obese (54% versus 28% of owners of dogs of normal weight), as stated above (Mason, 1970; Kronfeld, 1988) and fairly inactive. The owners of obese animals translate every appeal by the animal as being an appeal for food. These owners clearly have little concern for balancing the diet. Some of these aspects are very familiar to practitioners (Kienzle et al., 1998).

Saint Bernard puppies. Ad libitum feeding must be advised against as most dogs are unable to regulate their food consumption. It is better to feed puppies of a given litter separately during the weaning stage. (© Psaïla).

The data presented above may appear to be fairly discouraging at first glance and they do not clearly differentiate the simple correlation (between the weight of the owner and the weight of the dog) and the causes of obesity. They are however, very useful for developing methods of preventing and treating canine obesity. They help focus attention on environmental factors in the wider sense, at first glance external to the animal itself but of primary importance to the health of the animal.

The owners of obese dogs may also interpret bulimia as a sign of good health (Kronfeld, 1988) and excess weight as a sign of beauty in some breeds. Some owners also make the mistake of using foods as a palliative treatment to prevent unmonitored animals from becoming bored or destroying things. Lastly, for a dog living in a family environment, receiving food from children (as a reward or in a game) can become a bad habit. Multiple pet households may pose problems with respect to individual food consumption. Nevertheless, contrary to a predetermined idea, the frequency of obesity is greater in households with only one dog (Kienzle et al., 1998; Robertson, 2003).

In conclusion, it appears that the energy requirements of dogs are often incorrectly estimated and accordingly, in many situations energy intake can be excessive. It is up to the clinician to determine whether the obesity is primary or secondary so as to establish the subsequent treatment.

Individual rationing is necessary in a group of several dogs, even if they are of the same breed. (© Campus Royal Canin).

Pathology Associated with Obesity

Until the end of the 1980s there was little clinical data on the conditions associated with obesity in dogs. Some epidemiological studies conducted on humans have been extrapolated to the dog. However, the simple extrapolation of human data on diabetes mellitus, hypertension, etc. to dogs is not satisfactory. It is necessary to study clinical data on obese dogs. A survey is given in Table 6.

Reduction in Longevity

It has been clearly shown that obesity leads to a reduction in the longevity of dogs. Kealy et al. (2002) followed a group of 48 Labradors, half of which received limited quantities of food throughout their life. From the start a group of dogs were fed ad libitum with a growth food and a second group received 75% of the energy intake of the first group. The body weight of all the animals in the ad libitum group as well as the group that received a lower energy intake, increased until age 3 years and 4 months, reaching the average values 35 kg and 27 kg respectively. At this point two dietary modifications were introduced: a food with a lower energy concentration was given to all dogs and the intake of the ad libitum group was restricted (provision of a set quantity of food), while the second group continued to receive 25% less energy.

The modification of the experimental protocol induced a reduction in the body weight of all dogs, which was subsequently stabilized. At 5 years of age, the average difference between the two groups of dogs was 10 kg. At 8 years of age, the body index was 6.8/9 for the dogs eating the most food and 4.5/9 for the control group dogs (1: cachexia; 9 : massively obesity).

At age 12 years the average energy intake of the control group dogs and the dogs receiving 75% of intake were 1745 kcal and 1352 kcal (around 127 kcal/kg and 115 kcal/kg of body weight (BW) 0.75) respectively. The dogs of the second group weighed on average 26% less than those of the control group. The dietary restriction helped prolong longevity to 13 years rather than 11.2 years in the control group. The energy restriction helped slow down the development of chronic diseases and more specifically arthrosis. Furthermore, different metabolic parameters (insulin, glucose, blood lipids) were also favorably influenced in the dogs receiving 25% less energy. The above study is extremely important in terms of its contribution to science: it confirms an undeniable relationship between energy intake and longevity in dogs. It constitutes an argument against ad libitum feeding and provides valuable data showing the consequences of obesity for the development of osteoarticular diseases.

The positive effect of energy restriction on life expectancy has also been observed in humans. Individuals presenting with an average body mass index live longer than overweight individuals (Manson et al., 1987).

The food consumption of large-breed puppies must be monitored from a very early age. (© Renner).

Osteoarticular Conditions

Obesity predisposes dogs of all ages to osteoarticular pathology (Figure 2 and Chapter 11). Obesity associated with over-consumption of food in large-breed puppies during the growth phase leads to the development of various orthopedic complaints or exaccerbated hip dysplasia (Kealy et al., 1992). The symptoms of osteoarticular disease associated with obesity are generally observed after 6 months of age. In many cases, the lesions are irreversible.

Figure 2. Relationship between orthopedic pathology and obesity: spread of hospitalized dogs based on their body score: scale from 1 (cachexia) to 9 (proven obesity) (Lhoest et al., 2004).  

In the above study of Kealy et al., the radiological signs of osteoarthritis of the coxofemoral joint have been studied in Labradors from 4 months of age. They gradually became more common in the Labradors fed ad libitum compared with the dogs whose energy consumption was limited (52% versus 13% at 5 years of age). In the Labradors 8 years and over, the most common chronic disease was articular arthrosis affecting several joints (shoulder, elbow, hip, knee), with 90% of the Labradors afflicted (Kealy et al., 1997, 2000, 2002). The study reported that the arthrosis was more severe in the group initially fed ad libitum.

Other orthopedic conditions are common in obese dogs (Janicki & Sendecka, 1991). Torn cruciate ligaments and fractures of the humeral condyle were the subject of a wide epidemiological study on 854 Cocker Spaniels. The dogs afflicted with one of two orthopedic complaints presented a body weight greater than that of the healthy population. Interestingly, the reverse was true of those dogs with disc disease (Brown et al., 1996).

The dogs suffering from osteoarticular diseases are caught in a vicious circle: the animal tends to reduce its activity, which leads to overfeeding and overweight if the energy intake is not adjusted. Furthermore, the observation of an articular condition (a torn cruciate ligament, for example) is certainly a major indication of the need to reduce the dog’s weight, but it may also frustrate this process by making exercise impossible.

Hip dysplasia. Major dysplasic lesions at left with subluxation and arthrosis. Obesity is a cause or an aggravator of hip dysplasia. Energy restriction helps slow down the appearance and development of the different forms of osteoarthritis. (© HAW Hazewinkel).

Intolerance to Effort and Cardiorespiratory Complaints

The principal symptoms associated with obesity are intolerance to effort (De Rick & De Schepper, 1980) and respiratory problems (Ettinger, 1983). There is also a link between the frequency of tracheal collapse and obesity, although the correlation with other factors, such as breed, is greater (O'Brien et al., 1966; White & Williams, 1994).

A field study has shown that when weight loss is achieved, the owner inevitably observes a change in the dog's behavior, whereby the dog is more alert and playful (unpublished Royal Canin data 2001, obtained on 13 dogs suffering from obesity for over a year and followed for at least 10 months).

The weight increase in the dog is accompanied by an increase in heart rhythm, ventricle volume, blood pressure and plasma volume (Rocchini et al., 1987; Mizelle et al., 1994; Massabuau et al., 1997). The correlation between obesity and hypertension is however, controversial. There was a link between age and the increase in arterial blood pressure in dogs, but not between obesity and hypertension (Bodey & Michell, 1996). Dogs have however been used in experiments to study the pathogenesis of hypertension induced by the increase in body weight and the associated insulin resistance (Verwaerde et al., 1997; Truett et al., 1998).

The frequency of cardiovascular diseases increases with obesity. Some clinical studies report diseases including thrombosis of the portal vein (Van Winkle & Bruce, 1993), hypoxia of the myocardium (Baba & Arakana, 1984) and valvular endocarditis (Valtonen & Oksanen, 1972; Edney & Smith, 1986).

The cardiovascular effects described above are also significant for nephrologists (Alonso-Galicia et al., 1995; Joles, 1998). Can hypertension lead to modifications in renal function over time? A clinical study has shown that in dogs overfed for six months with a diet rich in animal fat, weight gain (58% heavier than the control group dogs) was accompanied by an increase in the weight of the kidneys (31% heavier), an increase in arterial pressure, glomerular filtration rate, renal blood flow and various renal histological lesions. The authors concluded that the lesions and the anomalies observed could be more severe in the case of prolonged obesity (Henegar et al., 2001). This study also suggested that the negative effects may be due not only to the influence of dietary fats but also to the composition of the fat.

English Bulldog. Obese dogs are more prone to heatstroke when the ambient temperature rises compared with dogs that are not obese. (© Lanceau).

Diabetes Mellitus

Diabetic dogs may present with hyperphagia leading to weight gain at an early stage. The correlations between obesity and glucose metabolism are complex, but it is clear that obesity leads to profound changes in the metabolism of glucose and the secretion of insulin (Mattheeuws et al., 1984a, b). It has been shown that the secretion of insulin, insulinemia and glucose intolerance increase proportionally to the degree of obesity. These changes are caused by a state of insulin resistance, one of the elements of which is chronic inflammation (Festa et al., 2001). The model of the dog overfed with a diet rich in fats has also been widely used to study the syndrome of insulin resistance. Indeed, when the dog's obesity is induced by feeding ad libitum with a diet rich in fats, insulin resistance will gradually develop in relation to an increase in adiposity (Rocchini et al., 1987; Bailhache et al., 2003a; Kim et al., 2003) and an increase in the production of adipocyte cytokines (Gayet et al., 2002, 2003b, 2004a, b; Jeusette et al., 2004b).

It has not been clearly established that obesity is a risk factor in the development of diabetes mellitus in dogs. Nevertheless, the current increase in the incidence of diabetes mellitus in dogs could lead to that supposition (Hoenig, 2002). (© Faculty of veterinarian of Liège).

Reduction in Immunity

Obese animals or animals fed a diet with a high fat content are less resistant to infection than animals fed a balanced diet (Newberne, 1966, 1973; Williams & Newberne, 1971; Fiser et al., 1972).

Hyperlipidemia and Dyslipidemia

According to Joshua (1970), fatty infiltration of the liver can be observed in obese dogs. An epidemiological study also shows that obesity increases the risk of acute hemorrhagic pancreatitis (Hess et al., 1999). Some results show profound disruptions to lipid metabolism. Obese dogs present with increased lipid plasma concentrations - cholesterol, triglycerides and phospholipids - without exceeding the reference values for these parameters (Chikamune et al. 1995; Bailhache et al., 2003b; Diez et al., 2004). An increase in the contents of non-esterified fatty acids and lipoprotein modifications (increase in triglycerides in the VLDL and HDL, reduction in HDL cholesterol and increase in VLDL cholesterol) (Bailhache et al., 2003a, b) have also been observed. The consequences of these well-known modifications in humans must however, still be evaluated in dogs.

Incontinence and Urinary Calculi

The hypothesis that there is a correlation between obesity and some forms of urinary incontinence, primarily in neutered bitches, has been suggested, but remains controversial (Gregory, 1994). Some bitches become incontinent after becoming obese and weight loss helps solve the problem. In some cases, incontinence will be observed again in bitches that have regained weight after losing it. One hypothesis that has been advanced is that the presence of retroperitoneal fat can exercise mechanical effects on the bitch's urinary system (Holt, 1987). The fact that neutered females are twice as likely to be obese compared with intact females must also be taken into account. This could explain the correlation between urinary incontinence and neutering. The debate is far from over.

Overweight dogs will also be more likely to develop urinary calcium oxalate calculi (Lekcharoensuk et al., 2000).

Reproduction Problems

The correlation between obesity and reproductive problems is not clear, although it is accepted that excess fat may lead to dystocia (Edney & Smith, 1986; Sonnenschein et al., 1991; Glickman et al., 1995).

German Shepherd bitch and puppies. In humans, it has been shown that obesity reduces fertility (Pasquali et al., 2003). This may also be the case in dogs. (© Renner).

Cancers

The correlation between obesity and some cancers (breast, uterus, colon and prostate) is well established in humans (National Institute of Health, 1998). Conversely, the lack of clinical data means such a link cannot be made in dogs with respect to anything but mammary tumors.

The first data were published in 1991. According to Sonnenschein et al., obesity or the consumption of a diet high in fat one year prior to the diagnosis does not increase the risk of mammary cancer in adult bitches, neutered or intact. These results have been contradicted by Perez Alenza et al. (1998, 2000).

On the other hand, the risk in neutered females was reduced in individuals that were slim between 9 and 12 months (Sonnenschein et al., 1991) and increased in females at the age of one year (Perez Alenza, 1998, 2000). On the whole, the authors conclude that the condition of obesity in juvenile animals certainly plays a role in the predisposition to mammary tumors in adulthood.

A retrospective study has not confirmed these results (Philibert et al., 2003). First of all, it was not possible to analyze the effect of early onset obesity on the development of mammary tumors. Neither did the authors report any correlation between obesity and the development of tumors, or between obesity and the period of survival (10 months for obese bitches versus 14 months for others).

Dermatological Disease

The many reviews that deal with canine obesity often mention that skin problems are more common in obese dogs than in healthy dogs. Paradoxically, to our knowledge there have been few studies to demonstrate this fact. In a clinical study on 29 dogs suffering from dermatitis due to Malassezia pachydermatis, obesity was identified as a significant risk factor for the development of this dermatitis (Pak-Son et al., 1999).

According to Edney & Smith (1986), the correlation between skin problems and obesity is not clear.

Exploratory Techniques

It is more difficult to use some exploratory techniques on obese dogs than on healthy dogs: auscultation, palpation or radiography are complicated by the excess subcutaneous or abdominal fat (Joshua, 1970).

Surgical Inconvenience

The risks associated with anesthesia are greater in obese dogs, but vary depending on the type of anesthetic used. The major risks are overdosing and lengthening of the period of recovery due to the storage of lipid soluble anesthetics in body fat. The other risks are associated with concurrent diseases that are common in obese patients, including circulatory, respiratory and hepatic problems (Clutton, 1988). In a controlled study on the surgical times in bitches undergoing ovariectomy, the surgical time was significantly - on average 30% - longer among obese bitches (Van Goethem et al., 2003).

In obese humans the surgical risk is increased due to various anomalies such as disruptions of the respiratory system, (reduction in respiratory capacity, hypoventilation), circulatory system (hypertension and cardiomegaly) or other functions (difficulties inserting tubes or maintaining water balance). Post-surgical complications are also more common in obese patients (Fisher et al., 1975).

Reversibility of Problems

• Intolerance to effort, inactivity, orthopedic and respiratory problems reported by owners are generally attenuated or may even disappear completely after weight loss (Gentry, 1993; Diez et al., 2002, 2004).
• The same is true for some heart rhythm problems (Baba & Arakana, 1984).
• Urinary incontinence may also be reduced or disappear completely after a diet (Holt, 1987).
• Recent studies have shown the reversibility of the main metabolic complaints, particularly insulin resistance and disruptions to lipid metabolism (Gayet et al., 2003a, 2004a, b; Jeusette et al., 2004b).

Modifications in Thyroid Function

Thyroid function has been explored in obese dogs in comparison to a group of healthy dogs and in the course of a weight loss protocol. The concentrations of some thyroid hormones were higher in obese dogs, and they decline during the weight loss protocol. The authors concluded that obesity and the restriction of energy alter thyroid function but that these modifications should not alter the interpretation of clinical trials (Daminet et al., 2003).

Pathophysiology of Obesity

In simple terms, obesity is the consequence of energy imbalance where intake exceeds expenditure for a variable period of time, which leads to a positive balance. There are a huge number of factors that can cause this situation and the interaction between these factors rather than the action of any one of them is felt to be responsible for obesity.

Energy Balance

The Principles of Energy Balance

The fundamental principal of energy balance is:

Modifications of reserves = energy intake - energy expenditure

A positive energy balance is the consequence of energy intake exceeding expenditure, and conversely, the balance is negative when expenditure exceeds intake. In normal conditions the energy balance oscillates meal after meal, day after day, week after week, without changing body weight and energy reserves in the long term. Many physiological mechanisms play a role in adapting intake to expenditure and expenditure to intake so as to maintain a stable body weight in the long term. If the energy balance is positive, expenditure increases (pointless cycles, uncoupling proteins, etc) and conversely, when the balance is negative, the body tends to reduce its expenditure (which contributes to the resistance to weight loss).

Energy Intake

The total energy intake provides all the food ingested, digested and metabolized by the body. Table 7 shows the energy intake via different nutrients that provide energy. The coefficients used are derived from Atwater's and involve some risk of error, since they take account of only average digestibility. Fat provide more energy per unit of weight than digestible carbohydrate or protein. In carnivores, dietary fiber is not very digestible and energy contribution is negligible. It should however be noted that an energy value of 1 - 2 kcal/g is attributed to digestible fiber in humans. In dogs, some soluble fiber is completely digested (Diez et al., 1998) and acetate can contribute to 8% energy metabolism in the dog (Pouteau et al., 1998).

Energy Expenditure

The second element in the equation is energy expenditure, which is split into three parts:

• Basal metabolic rate (BMR)
• Postprandial thermogenesis (production of heat subsequent to the meal)
• Physical activity

The most commonly used equation to calculate the theoretical maintenance energy requirement is:                 132 kcal per kg of metabolic weight (MW)*                                                                 * Where MW = (body weight) 0.73                                                                         This exponent is often rounded up to 0.75 to facilitate calculations, but the original value is 0.73.

In sedentary human adults, basal metabolic rate, postprandial thermogenesis and physical activity represent 60%, 10% and 30% of energy expenditure respectively (WHO, 1997). The contribution of each of these factors varies significantly however, depending on the regularity and intensity of the physical activity, which is the key variable in expenditure. Basal metabolic rate appears, on the other hand, to be a stable individual factor, that has a major impact on the amount of muscle mass of the organism (90 - 95% of energy expenditure of basal metabolism versus 5 - 10% for fat mass).

In dogs, the basal metabolic rate also represents 55 - 70% of the total expenditure (NRC, 2006), but differences have been observed between breeds. By way of example, the Labrador has a lower basal metabolism rate than the Great Dane or the Spaniel. The basal metabolism rate declines with age in dogs (Speakman et al., 2003). It is recommended to reduce the energy intake by 10-15% from age 7, while adjusting the diet depending on the physical condition of the individual. On the other hand, a low calorie diet is not always justified for all old dogs.

The necessary balance between intake and expenditure is the crux of the problem in dogs in general and in obese dogs in particular. The difficulties of estimating energy expenditure (requirements) are multiple.

First of all, the great diversity of the canine species: body weight varies between the extremes of 1 kg and over 100 kg. It is also easy to understand the difficulty of estimating the energy requirement of all the dogs with a simple equation. The average equation is 132 kcal/kg BW 0.75 (NRC, 1974). An initial approach proposed was to group breeds on the basis of their weight and size: small, medium, large and giant.

Dogs of comparable weight and size in the same category may present very different energy requirements. The differences may be due to the thickness of the skin, the body composition (lean mass/fat mass ratio) or to the type of selection used (originally a working dog, then selected on beauty criteria and having the function of a companion animal). The body composition is extremely important: the dogs that present a great muscle mass expend more energy and are less susceptible to obesity than fat dogs.

Breed can have a strong influence on energy requirement, even between two dogs of similar weight in similar environmental conditions. When it comes to theoretical rationing (NRC 1974), it is wise to reduce the Newfoundland's ration by around 10%. The ration of a Great Dane on the other hand must often be increased by 40% to maintain the dog's weight. (© Royal Canin/J.-P. Lenfant - © Royal Canin/Renner).

Besides breed, individual factors, both genetic and otherwise, also generate great diversity in energy requirements. In dogs of the same breed of comparable weight, the males are generally a little less fat than the females and so their expenditure is greater (on the order of 10%) however, this is a controversial point (Kienzle & Rainbird, 1991).

As observed above, neutering will reduce expenditure (to the order of 20 - 30%) (Figure 3). Aging of the animal is an example of a physiological condition that may reduce energy expenditure by reducing the basal metabolic rate. Moreover, body composition changes during aging. The fat mass tends to increase at the expense of the muscle mass.

Energy expenditure associated with physical activity has not been quantified in dogs. From a practical perspective, it is not possible to say how many kilocalories correspond to an hour's walking, hunting or racing.

Figure 3. Development of average weight of beagle bitches after neutering (Jeusette et al., 2004a).

In a neutral thermal environment, the energy expenditure associated with the thermoregulation of dogs living indoors is low. The seasonal effect is accordingly negligible. On the other hand, to maintain thermoregulation the energy expenditure of dogs kept in an outdoor kennel increases when the ambient temperature falls. In the literature however, the quantification of supplementary energy expenditure is controversial. In the German Shepherd for example, a variation of 1°C is accompanied by a variation in the energy requirement of the order of 1% (Manner, 1991). Other data show an increase from 2.3 up to 3.8% per degree below the thermal neutrality zone (NRC, 2006).

To conclude, the estimation of the energy requirement is not easy in dogs. While there is a large quantity of data, the data is fragmented and it is difficult to generalize. In practical terms, monitoring weight and knowing what the dog needs to consume to maintain a constant weight are the most important pieces of energy-requirement-based information for any given individual.

The Physiological Regulation of Body Weight

In their original habitat, wild canids are generally active and in an environment with abundant food it is extremely rare for adult animals to be obese. The biological mechanisms regulating body weight are present throughout the animal kingdom and appear to be fairly efficient in combating underconsumption.

In domesticated animals, the environmental pressure makes the mechanisms likely to manage overconsumption more useful. The hormonal regulators of appetite, food consumption and energy expenditure are becoming more and more well-known, particularly leptin, ghrelin and adiponectin.

Leptin is a cytokine produced and secreted by the fat cells. It acts as an energy balance modulator signal, both centrally (on the hypothalamus) and peripherally (liver, pancreas, etc). Leptin accordingly appears to play a key role in the regulation of food consumption. At the time of its discovery leptin was presented as a miracle treatment for obesity as leptin injections in obese and healthy mice induced significant body weight loss without any apparent side effects. Nevertheless, it has been shown that obese humans and dogs (Ishioka et al., 2002; Gayet et al., 2003a; Jeusette et al., 2003, 2004b) do not suffer from leptin deficiency. On the contrary, in these two species, leptin is produced in proportion to the quantity of fat cells such that the rates of plasma leptin are higher in obese individuals than in healthy individuals. Leptin increases energy expenditure in healthy individuals but in obese individuals the situation is less clear due to a resistance phenomenon.

On the other hand, it appears that insulin and many other mediators also play a role in regulating leptin (Lonnqvist et al., 1999). Clinical studies conducted on humans tend to show that the blood concentration of leptin is dependent on the secretion of insulin, the composition of the food and exercise (Koutsari et al., 2003). From a practical perspective, the main point to remember with respect to obesity is that leptin reduces the appetite. Leptinemia increases in dogs during periods of weight gain (Gayet et al., 2003a, 2004b; Jeusette et al., 2004b).

Ghrelin (GH releasing hormone) has been identified by Kojima et al. (1999). It stimulates the secretion of growth hormone (GH) and increases food ingestion in humans and rodents. We have observed that the plasma concentration of ghrelin is lower in obese dogs than it is in healthy animals (Jeusette et al., 2003, 2004b).

Adiponectin is a cytokine secreted exclusively by the adipose tissue. It affects carbohydrate homeostasis, sensitivity to insulin and probably energy homeostasis. It will act in synergy with leptin (Yamauchi et al., 2001). Its expression is reduced in obese and diabetic mice (Hu et al., 1996). It is also reduced by half in obese dogs compared with healthy dogs (Gayet et al., 2004b).

Additional regulatory factors include TNF-α (tumor necrosis factor). This cytokine was originally identified as a pro-inflammatory molecule that participates in anorexia and cancer cachexia. It has been found in particularly high quantities in the adipose tissue of obese animals and patients. The expression and the concentrations of TNF-α are positively correlated to the degree of obesity and the resistance to insulin (Hotamisligil et al., 1995) as shown in dogs (Gayet et al., 2004a).

Besides the mechanisms mentioned above, the activity of uncoupling proteins (UCP) deserves emphasis. These proteins belong to a family of transporters in the internal membrane of the mitochondria, which uncouple the respiration of ATP synthesis by dispersing the mitochondrial proton gradient. The activity of these proteins varies according to thermoregulation and postprandial thermogenesis. The expression of UCP-1 is greatly reduced in the adipose tissue of obese, insulin-resistant dogs (Leray et al., 2003).

To conclude, it appears that many factors involved in the development of obesity in humans and rodents have also been identified in dogs, whether they are factors limiting appetite or factors increasing expenditure.

• Leptin is a protein that increases energy expenditure in healthy individuals.
• Ghrelin is an orexigenic hormone principally secreted by the stomach and the duodenum.
• Adiponectins acts in synergy with leptin. it is secreted by adipose tissue.

The Dynamics of Weight Gain

Despite these regulator mechanisms, a positive energy balance can induce weight gain if it continues for a long enough period of time. There is controversy as to the period of imbalance (energy intake greater than expenditure). In humans the perceived hypothesis is that obesity is established slowly, following a prolonged imbalance (several years) that does not have to be sizeable. Clinicians distinguish three phases:

• A pre-obesity static phase during which the individual's energy intake is increased but its weight remains the same
• A dynamic phase, during which the individual gains weight, primarily increasing its fat mass but also its non-fat mass, if only by a low increase in the blood volume
• A static phase during which the balance between intake and expenditure is re-established due to the reduction in food consumption. In this phase, the weight is extremely high, but the basal metabolic rate is relatively low. In this new state of balance obesity is usually considered to be morbid (WHO, 1997).

From a certain phase of obesity the food consumption may decline without the dog losing weight, because the basal metabolic rate is relatively low. (© Faculty of veterinarian of Liège).

These data can almost be extrapolated to companion dogs, after a few adjustments. With respect to the period of imbalance, the data must be adapted to the life expectancy of the dog and practical observations must be taken into account. Weight gain can occur rapidly, in a few weeks or months. Rapid weight gain may occur in the weeks following the neutering of bitches. Ad libitum feeding during growth may induce major weight gain in puppies aged 8 months.

The pre-obesity phase has not been described in dogs. The dynamic and static phases on the other hand have been well documented (Figure 4). The dynamic phase may be linear or segmented. Weight is stabilized during the static phase; appetite may be normal or diminished. This explains why it is common to see obese dogs in the clinic that "do not eat a lot". What is clear is that the energy intake was greater than the dog's needs at a given moment in its life - sometimes several years earlier. But once these animals have been stabilized, their energy requirements are low and physical activity is often very limited, the more so at this stage.

Quality of Dietary Intake

Various studies conducted on humans and laboratory animals show that dietary factors, particularly energy intake and fat content, are directly correlated to obesity.

Figure 4. Weight development of three beagle dogs fed with a diet rich in fats.  

Energy Intake and Macro-nutrients

For dogs, the calculation of the energy provided by a food is based on its chemical composition. Fats are the nutrients that contain the most energy. The over-consumption of fatty food is accordingly an essential factor in the genesis of obesity. Fat is added to food to both increase palatability and the energy density of the diet.

In terms of metabolizable energy, the protein and carbohydrate intake is equal. When the net energy intake is calculated however, the utilization of proteins for energy is lower (Table 7) (Rubner, 1902). This is one of the reasons, other than the specific effect of some amino acids (lysine, phenylalanine, leucine), for proteins having more of a satisfying effect than carbohydrates. The carnivorous nature of dogs may explain to some degree the resistance of wild canids to obesity.

Digestible carbohydrates clearly provide the same quantity of energy, but they induce different metabolic effects, particularly on the secretion of insulin. This point is addressed in more detail below.

From a theoretical perspective, it can be accepted that the mathematical adaptation of energy intake to energy expenditure suffices to prevent canine obesity. But this will be disregarded in some cases, because it is based on "metabolizable energy". The simple modification of the food's chemical composition - without modifying the total energy intake - can lead to changes to the body composition and the basal metabolic rate. This point has been proven in dogs and is well established in human nutrition (Bouché et al., 2002).

Diagnosis and Evaluation of Obesity

One of the major tasks facing the clinician is to evaluate how obese the animal is as in many cases the optimal body weight is unknown. In human medicine, it is easy to calculate an optimal weight range based on size, using the BMI (body mass index), which is the ratio of height to weight. In human medicine, doctors have a BMI reference table they can consult. These tables are not available for domesticated carnivores. Various attempts at morphometric measurement have proved inconclusive due to the great diversity of canines breeds. Other less standardized tools have accordingly been proposed for veterinary medicine.

Body Weight

The simplest method is the body weight reference. It is easy to weigh a dog, but weight alone is not enough to evaluate obesity. Without an indication of the dog's ideal weight this data is of little use. While it is easy to use breed standards as a reference for purebred dogs, this method is not completely satisfactory, because the animal's body weight can vary quite significantly depending on its stature (Table 8).

When visiting the veterinarian it is important that the dog is weighed and its medical file is updated accordingly. The ideal weight must be identified or estimated to ascertain an obese animal's diet. This is the most important factor in formulating a diet that will enable the dog to lose weight.

Morphometric Measures

The combination of data on stature and on body weight introduces the concept of morphometric techniques to evaluate body composition. Morphometry measures the outer form, evaluates certain body regions and how their dimensions change and shows their relationship to modifications in body composition. The morphometric techniques used on dogs are body condition scoring and techniques that combine diverse body parameter measurements (length and circumference of various parts of the body).

A massive deposit of adipose tissue on the spinal column and the base of the tail are among the criteria used to identify obesity.

King Charles Spaniel puppies (© Renner).

English Bulldog puppies (© Lanceau).

Labrador puppies (© Hermeline/Doxicat).

Bernese Mountain Dog puppies (© Renner King).

The body index is a semi-quantitative subjective evaluation method combining the evaluation of visible characteristics and palpation of certain regions of the body. This evaluation is conducted in accordance with simple criteria: the size and location of major adipose deposits, the visible and invisible skeletal structure, and the silhouette of the animal.

Several types of index have been proposed:
- 3 grades: 1 = slim, 2 = optimal, 3 = excessive
- 5 grades: 1 = gaunt, 2 = slim, 3 = optimal, 4 = overweight, 5 = obese (Edney & Smith, 1986) (Table 9)
- and even 9 grades: 1-4 = emaciated to slim; 5 = optimal; 6-9 increasingly overweight (Laflamme, 1993; Laflamme et al., 1994a).

The animals presenting an average index corresponding to an optimal weight have a fat mass of around 13%. When a 9-grade body index is used, every grade in the index represents a 9%-increase in fat mass (Mawby et al., 2000). As a consequence, an animal presenting a body index of 9, which corresponds to the qualification "morbid obesity", has a fatty mass of over 40%. The advantage of these index systems is that they can be easily applied by the clinician and that they do not apply exclusively to the diagnosis of obesity but also to its active prevention. It is easy to weigh the animal during a routine consultation and find the value in the index.

The transition from a 5-grade scale to a 9-grade scale is easily accomplished by applying intermediate graduations in the 5-grade scale. Here, the body index of this Beagle bitch can be 4.5/5 or 8/9. (© I. Jeusette)

The measures of the various circumferences - thoracic and pelvic for example - and their use in the equation systems do not enable any appreciation of the fat mass due to morphological differences between individuals. They are however a good way of monitoring the weight loss in a given dog. The various body measures do require experience from the practitioner, as well as the cooperation of the animal (Burkholder, 2000).

Ultrasound Measurements

Ultrasound has been used to measure the thickness of the subcutaneous fat layer in dogs (Anderson & Corbin, 1982; Morooka et al., 2001). Combining this technology with others may be a good way to localize the main fat deposits and understand the mechanisms that drive the development of obesity (Morooka et al., 2001). Furthermore, this technology is relatively simple and non-invasive; it is useful in the clinic provided good equipment is available. Its field of application is narrow however - it can only be used to evaluate subcutaneous fat. Several body regions have been tested, including the middle of the lumbar region and the L6 and L7 lumber or S1 sacral apophyses. The problem is reproductibility: it is necessary to shave the hair, standardize the animal's position and the pressure of the sound probe, and use objective benchmarks. The use of ultrasound in two dimensions helps improve the technique and obtain more precise figures (Morooka et al., 2001).

DEXA

The use of DEXA (Dual Energy X-ray Absorptiometry) (Munday et al., 1994) is a way of differentiating between the nature and the quantity of each tissue in the parts of the organism subjected to examination and to monitor the development of the dog's body composition during the period of weight loss. This examination requires that the animal is anesthetized for the procedure. The results relate to the bone mineral mass, adipose tissue and lean mass of the organism (Figure 5).

Figure 5. Images (obtained by DEXA) of a neutered male labrador age 4.5 years, before (T0) and 5 months after (T+5) the introduction of a low calorie diet. (© Alex German).  

Heavy Isotopes

Body water is primarily stored in lean tissue; as a consequence, it is an indirect measure of lean mass. The quantity of total body water can be estimated by establishing the dilution of deuterium oxide (D2O) or of water enriched with O18. The fat mass and its percentage can be calculated on the basis of the difference. Deuterium and O18 are two excellent non-radioactive and non-toxic tracers at low doses. This method requires a blood sample before the subcutaneous injection of the tracer and a second blood sample four or five hours post injection. It can be used in practice to estimate the percentage of fat tissue in obese dogs, provided there is access to mass spectrometry. This non-invasive method has been validated in dogs (Pouteau et al., 1998; Son et al., 1998).

Bioelectric Impedance

In humans, the measurement of bioelectric impedance is a fast and simple non-invasive method for studying the body composition that is both portable and reproducible. The method has been tested on cats and dogs (Elliott et al., 2002a, 2002b).

These three methods - which have only recently been used on carnivores - produce well correlated results (Son et al., 1998). They are certainly more applicable to research settings than to clinical situations, but nevertheless, they do open interesting perspectives for comparing the efficacy of different low energy foods available in the marketplace (Diez et al., 2002).