Introduction

Dogs show unique diversity in morphology and body size, with small to giant sized breeds1,2. Hence, the variation in body weight (BW) between breeds is extensive, and at the same time BW can also vary considerably within breeds3.

The adult BW of a dog is reached at different ages between breeds and depends on both genetic and environmental factors2,3,4,5. High BW in dogs, due to large size or excess amount of body fat, is a risk factor for several developmental and metabolic alterations6,7,8,9 and degenerative joint diseases, such as hip dysplasia and osteoarthritis, and can negatively affect quality of life and lifespan10,11,12,13,14,15,16. High BW due to excess body fat is an escalating problem in dogs as well as in humans17, and studies have reported 20–59% of dogs being overweight or obese18,19,20,21,22,23,24,25.

The basis for indicated breed BWs in breed standards are often unknown. Given the connection between BW and canine health, scientifically evaluated breed BWs of young adult dogs of different breeds are warranted. In addition, it is currently unknown whether breed BW of young adult dogs has been stable over time or been subjected to change.

In the official screening program for hip dysplasia, run by the Swedish Kennel Club (SKC), recording of the dog’s BW at time of screening has been required since 2005. The program is available for all breeds for an official screening result from an age of 12 months (or 18 months for certain large breeds) and with mandatory registration and publication of all results in the SKC database. With approximately 70% of dogs in Sweden being registered in SKC, the SKC database holds an extensive amount of individual observations on BW in young adult dogs of different breeds26, which are explored in the present study. The objectives of this study were to: (1) determine mean breed BW in young adult dogs of different breeds and sex in Sweden, and (2) study changes in breed BW over a 10-year period in young adult male and female dogs.

Materials and methods

For this observational study, a data file with hip radiographic screening results on dogs born during the years 2005–2015 was acquired from SKC in October 2018. The file held screening data from 2005 and onwards and contained information on the individual dogs, including registration number, breed, sex, date of birth, screening date and BW at time of screening.

Data management

Cleaning of the data file was performed as follows: individuals without a Swedish registration number or with results from a former grading system were excluded from the file. Screening posts performed prior to the officially accepted age (12 months, or 18 months for certain large breeds) were excluded as were individuals with missing BW observations. Individuals with biologically implausible registered BWs (i.e. biologically impossible or extremely unlikely BWs for individual dogs, based on our understanding of biologically possible BWs for each specific breed)27 were excluded.

Screening results from 2006 and 2017/2018 were excluded, because the relative proportion of younger and older individuals, respectively, was different from the rest of the material. If an individual dog had more than one screening result, all but the first registered result were excluded. Individuals with screening results at an age of 12–24 months, or 18–30 months for certain large breeds (Tables 1 and Table S1), in breeds with at least 15 observations per year during the 10 years of screening (2007–2016), were included in the statistical analyses.

Table 1 Body weight (kg) per breed and sex for dogs screened during 2007–2016.
Full size table

After cleaning of the original data file (consisting of 183 252 records on dogs of 304 breeds), and applying the inclusion criteria, 114 568 dogs (64%) with individual BW observations from 72 breeds were included. The fraction of dogs participating in screening, of the registered dogs in the included breeds, ranged from roughly 5 to 72%, and was above 30% in 54 of the 72 breeds28.

Statistical analysis

Commercially available software (SAS 9.4, SAS Institute Inc., Cary, NC, USA) was used for data management and statistical analyses. General linear models, using PROC GLM, were used to evaluate BW at screening and the change in breed BW over time, with BW considered a continuous variable with potentially normal distribution.

The analyses were performed separately for each breed. Year of screening was treated as a continuous variable, coded 1–10, and the other variables in the models were sex (male/female), age at screening and the interaction between sex and age at screening. To allow for non-linear associations between BW and age at screening, we also included a centered and squared term of age at screening, and the corresponding interaction term, in the models. Possible interactions between sex and year of screening, indicating different trends by sex, were initially evaluated and, if significant, the effect of year of screening was estimated nested within sex. Least-square means (LSMeans), or marginal means, were used to illustrate the associations between BW and sex and were evaluated at a ‘mean term’ for age at screening (547 days old at screening) and year of screening (the 6th screening year). Relative change in breed BW (%) over the 10 years was calculated as ten times the estimated regression coefficients divided by the mean breed BW from the first screening year. As adjustment for multiple testing is not required or preferred in hypothesis-generating studies29,30, no adjustment was performed.

The residuals were inspected for signs of non-normal distribution and heteroscedasticity, but no deviations were found. Influential observations were identified using Cook’s distance, but no abnormalities were found (all D’s < 0.20). Statistical level of significance was set at P < 0.05 for all analyses.

Conference presentation

Presented in part at the 31st European College of Veterinary Internal Medicine—Companion Animal Congress, Online, September 2021.

Results

The number of individuals per breed are shown in Table S1. The sex distribution was equal in the majority of breeds, with slightly more female dogs than male dogs in total (53 and 47%, respectively). Three of the 72 breeds (4%), had > 10 000 BW observations, 30 breeds (42%), had > 1000 observations and 54 breeds (75%) had > 500 observations on BW. Sixty-five of the breeds included dogs screened at 12–24 months of age, and the remaining seven breeds included dogs screened at 18–30 months of age (Table S1).

Mean bodyweight per sex

Mean BWs per breed and sex are presented in Table 1. The BW ± SE ranged from 7.2 ± 0.1 kg in female Danish Swedish farm dogs to 69.2 ± 1.0 kg in male long-haired Saint Bernhard dogs. A significant breed BW difference between sexes was present in 54 (75%) of the breeds, indicated in the table by the P-value from estimates of the effect of sex on BW, with male dogs being more heavy in all breeds.

Change in bodyweight over time

In Table S2, the estimates of the effect of year on BW are shown for each breed. A significant change in breed BW over the 10 years was present in 33 (46%) of the 72 breeds (Table 2). Of these, five breeds showed an increase in BW over time, with no difference between sexes (estimates in kg/year ranging from 0.077 in Bernese Mountain Dogs to 0.313 in Gordon Setters), and 22 breeds showed a decrease in BW over time, with no difference between sexes (estimates (kg/year) ranging from − 0.042 in Golden Retrievers to − 0.223 in Newfoundlands).

Table 2 Regression coefficients for the association between year of screening and body weight in breeds with a significant change during the screening years, 2007–2016.
Full size table

Interaction between sex and year was present in six breeds with a change in BW, and, therefore, estimates were established for each sex in these breeds (Table 2). In Lagotto Romagnolos, BW had decreased in both sexes, but with a larger reduction in male (− 0.246 kg/year) than in female (− 0.145 kg/year) dogs. In Boxers and Spanish Waterdogs, only male dogs showed a decrease in BW (− 0.138 and − 0.154 kg/year respectively), while in White Swiss Shepherd Dog, only female dogs showed a decrease in BW (− 0.286 kg/year). In Shiba Inus and Belgian Shepherd Dogs (Malinois), male dogs showed an increase in BW (0.089 and 0.165 kg/year respectively), while female dogs showed a decrease (− 0.084 and − 0.094 kg/year, respectively).

Relative and numerical changes in breed BW over the 10 years are shown in Tables 3 and 4. In breeds with no difference between sexes, Gordon Setters showed the largest relative and numerical increase in BW during the 10 years, 14% (3.1 kg), and American Staffordshire terriers the greatest decrease; 8% (− 2.2 kg) (Table 3, Fig. 1a,b). In breeds with differences between sexes, the change in BW differed both in direction and magnitude (Table 4). By way of example, Lagotto Romagnolos showed a decrease of 15% (− 2.5 kg) in male and 10% (− 1.4 kg) in female dogs (Fig. 1c), while Shiba Inu male dogs showed an increase of 8% (0.9 kg) and female dogs a decrease of 9% (− 0.8 kg) (Fig. 1d).

Table 3 Ten-year change in body weight in breeds without significant sex difference (screening years 2007–2016), based on estimates in a general linear model and presented in % and kg.
Full size table
Table 4 Ten-year change in body weight in breeds with interaction between sex and year of screening (screening years 2007–2016), based on estimates in a general linear model and presented in % and kg.
Full size table
Figure 1
figure 1

(a-d) Observed mean body weight (BW) ± standard deviation in selected breeds with significant change in BW during the study period (2007–2016). Linear regression lines based on the observed mean values have been applied to the graphs.

Full size image

Discussion

In this large observational register-based study, we have estimated BW in young adult dogs of 72 different breeds in Sweden. In 33 of the breeds, a change in breed BW during the 10-year period was noticed, and, although not a prominent change for most, it was greater than 10% in some breeds.

The study population consisted of young adult dogs; 12–24 months old for most breeds and 18–30 months old for certain larger breeds. Adult canine BW at maturity depends on both genetic and environmental effects, and is reached at different ages. Factors including nutritional status and body condition, as well as breed, size and sex, have been associated with growth, time to maturity and adult BW2,3,4,5,31. The lowest age for an official hip screening result is 12 or 18 months. The age limit is set in agreement between international kennel club organizations and is based on age at maturity for each breed, with larger breeds reaching maturity later. In Sweden, dogs are traditionally examined at the allowed age for an official result or shortly thereafter, with results publically available in a database28. The inclusion criteria for age in the study were set with the intention to include dogs from young adult age at an age interval when the majority of dogs are screened.

The sex distribution was equal in most breeds in this study, which reflects the sex distribution in all dogs registered in SKC, as well as in the Swedish dog population32. The BW differed significantly between sexes in 54 of the 72 breeds, with male dogs being heavier. Studies have shown slower growth in male dogs of several breeds, but with higher BW at the end of the growth period than female dogs of the same breed3,5,31. The non-significant difference between sexes in some breeds in this study might depend on some participating individuals, especially in the larger breeds, not having reached fully adult BW at the time of screening3,4. Potentially, varying breeding priorities regarding sex characteristics between different breeds, might also affect the outcome. Included breeds were mostly medium to large sized breeds, which generally have a higher prevalence of hip dysplasia33,34 and therefore a larger number of screenings, compared to smaller sized breeds. Even if the screening program is available for all breeds, screening is a requirement before registration of offspring in certain breeds, while specific screening results are needed before registration in other breeds26. Therefore, the population coming to screening likely differs between breeds. There is a long tradition to screen dogs for HD in Sweden26, which reflects the individual dogs participating in the program. Being a tool for selection of breeding stock, many presumable breeding dogs attend the program. However, the majority of dogs attending the screening are not breeding animals28, but offspring participating for parental breeding evaluation or pet dogs participating for joint health evaluation, since the screening result indicates risk of future clinical signs and can thereby be of value for the individual owner15.

Adult BW is affected not only by body size, but also by body composition i.e. the proportion of lean mass and fat. Since such information was not available in the data set, it is not known how the BW reflects size or body composition of included dogs. Thus, despite the extensive study population it is difficult to evaluate whether dogs in this study population are representative, i.e. lighter or heavier than dogs in general, in Sweden as well as globally. The BWs given in the breed standards might not be valid references for comparison. They are not only commonly given as ranges or maximums, but might also be based on estimations and assumptions made decades ago, when the variation within breeds was greater than today.

Approximately 45% of the studied breeds showed a change in BW during the 10-year period. The change could be due to changes in body fat mass and/or in size. Regarding the health risks associated with high BW, due to excessive body fat or large size, the decrease in BW observed in several breeds, might imply a reduced risk of known health disadvantages, including orthopedic disorders10,11,13,14,15,16,35. The breeds with the most prominent decreases in BW, Lagotto Romagnolo, American Staffordshire Terriers, Staffordshire Bullterriers and Spanish Waterdogs, are primarily companion dogs in Sweden, but also to a great extent attend shows and competitions. These four breeds had a screening hip dysplasia prevalence of 30–65% during the studied years28, and the decrease in their breed BW could improve that situation. They are all popular breeds, and an increasing popularity is reflected in extensively rising numbers of registered individuals during the study years for Staffordshire Bullterriers and Spanish Waterdogs28. The rising popularity might have affected the population coming to screening and thus, the breed BW.

The five breeds that had a similar increase in BW in both sexes, i.e. Gordon Setter, Belgian Shepherd/Groenendael, Bullmastiff, German Wirehaired Pointer and Bernese Mountain Dog, were larger breeds of different types and uses. The largest increase was found in Gordon Setters. If this increase is due to an increased fat mass, it would reflect a one-step change in the semi-quantitative 9-point scale for body condition score (BCS)36. The Gordon Setter is a bird hunting dog, which has been bred into a field type and a show type. The show type is heavier than the field type, and a possible preference for the show type in later years may explain some of the observed change in breed BW in Gordon Setters. The Bullmastiff, one of the heaviest breeds in our study with a mean breed BW of 54.5 kg in male, and 46.0 kg in female dogs, respectively, also showed an increase in BW. The reported screening prevalence of hip dysplasia in the breed was > 50% during the study years in Sweden28. With BW being a risk factor for development of osteoarthritis in large-size breeds with hip dysplasia14,37, and the known risk of increased morbidity and mortality in these dogs15, the increase in breed BW in the Bullmastiff, might have negative impact on their overall health and lifespan, regardless of the underlying cause.

Internationally reported prevalence of overweight and obesity in dogs is high, and seems to be increasing18,19,20,21,22,24,38. Concerns for a general trend towards heavier show dogs, both older and younger, are being discussed23,38, even though recommended BW and height in breed standards have not changed for the last decades. In consistency with the internationally published figures, a recent study on Swedish show dogs found 32% of the dogs being overweight23. The observed decrease in mean BW in several breeds in our study might as previously mentioned, relate to either changes in size or body condition. One potential explanation for the decrease could be a change in dog owner’s view of normal body condition in young, active dogs of certain breeds during the 10-year period. It is, however, important to be aware that the study population itself and possible changes in its composition, might impact the results. Data was collected in conjunction with a screening program and the dogs were presumably healthy, young, adult dogs, in which overweight due to excess of body fat tends to be less common than in middle-aged and older dogs18,19,21,25,39. The assumption of dogs being presumably healthy is based on the intention of the screening program, i.e. being a tool for breeding evaluation and for evaluation of future joint health. The number of neutered dogs in the study population is most probably low, based on the participating potential breeding animals and their offspring, and the tradition in Sweden not to neuter young dogs extensively without medical reason. The effect of neutering on BW in this study could therefore be considered to be low, in consistency with a previous study on BW in young dogs3. Changes in breed popularity and usage are factors that might affect our results. Certain breeds containing subgroups with different phenotypes were included as one in the dataset. The Labrador Retriever, for example, showed a decrease in breed BW, despite being a breed with reported high prevalence of overweight25,40,41, and an identified gene mutation associated with adiposity and food motivation42. The division into two almost different breed types with the leaner hunting type increasing in popularity, might account for the unexpected decrease in BW in this breed.

There are some limitations to this study. The study population consists of dogs registered in SKC and participating in a screening program, which might lead to selection bias, and affect the representativeness for the general dog population in Sweden. The breed representation does not cover all breeds due to inclusion criteria on number of observations, and smaller sized breeds are therefore underrepresented. Our criteria also exclude dogs older than 2.5 years of age. However, around 70% of the total dog population in Sweden are registered in SKC, and the proportion of dogs participating in screening was above 30% in 54 of the included breeds, which is in agreement with the fraction screened in many breeds in SKC26. Thus, based on the above, and considering the large data set, we assess that our study population represents the included breeds and age groups relatively well. However, the rules regarding registration of dogs in SKC could also affect the results. For example, breed variations and subgroups are present in some dog breeds, based on e.g. size or coat color and texture, and therefore registered separately in SKC, while other breeds with obvious division into phenotypically different subgroups register as one. Rules for registration in SKC and for screening might have changed in certain breeds during the study period, which could affect the results. Lacking information about the body condition of the dogs, it is unknown whether the population described can be considered ideal weight or not, and what observed breed BW changes depend on. Finally, being an observational study, results of changes in BW over time need to be interpreted with caution, especially for breeds in which changes were small, and should be confirmed in further studies.

In summary, this study provides BW information for a large number of breeds, based on data from young adult dogs in Sweden, mandatorily registered in conjunction with a hip screening program. These breed BWs could proposedly be used as a guide, complementing other sources on adult breed BW. Over the 10-year period, a change in BW was present in around 45% of the breeds with a change greater than 10% in some. Lacking details about body condition, and thereby knowledge of what constitutes the BW, this could be a consequence of changes either in breed-related size or in body fat mass. Still, the change in BW in certain breeds might have an impact on the overall health within those breeds. Future studies with simultaneous evaluation of BW and body condition over time are warranted and might further elucidate this matter.