Methods and mortality results of a health survey of purebred dogs in the UK

Abstract

To collect information on the cause of death and longevity of dogs owned by members of the numerically largest breed clubs of 169 UK Kennel Club-recognised breeds.
A cross-sectional study was carried out. Approximately 58,363 questionnaires were sent out to breed club members in 2004 (nine clubs failed to report the exact number of questionnaires sent out). Owners reported age at death and cause(s) of death for all dogs that had died within the previous 10 years.
A total of 13,741 questionnaires (24% response rate) containing information on 15,881 deaths were included in the analysis. Breed-specific response rates ranged from 64·7 to 4·5%. The median age at death was 11 years and 3 months (minimum=2 months, maximum=23 years and 5 months) and it varied by breed. The most common causes of death were cancer (n=4282, 27%), "old age" (n=2830, 18%) and cardiac conditions (n=1770, 11%). Clinical
This survey shows breed differences in lifespan and causes of death, and the results support previous evidence that smaller breeds tend to have longer lifespan compared with larger breeds. Although many of the breeds in the study may not be representative of the general pedigree dog population in the UK, the results do contribute to the limited information currently available.

Full text

PAPER
512 Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association
Veterinary Epidemiology Consultant,
PO Box 80, Bury-St Edmunds, Suf folk,
IP28 9BF
*Animal Health Trust, Kentford, Suffolk,
CB8 8UU
†Kennel Club, 1-5 Clarges Street, Piccadilly,
London W1J 8AB
‡Cambridge Infectious Diseases Consortium,
Department of Veterinar y Medicine, University
of Cambridge, Madingley Road, Cambridge
CB3 0ES
V. J. ADAMS, K. M. EVANS*,
J. SAMPSON† AND J. L. N. WOOD‡
Journal of Small Animal Practice (2010)
51, 512–524
DOI: 10.1111/j.1748-5827.2010.00974.x
Accepted: 15 June 2010
Methods and mortality results
of a health survey of purebred dogs
in the UK
INTRODUCTION
Baseline data on canine mortality provide
an important frame of reference for future
research into the mechanisms of disease.
The wide variety of pedigree dogs that
exists today underlies the need for breed-
specific information. Recently developed
genetic and molecular techniques that are
available to study the causes of death and
disease require knowledge of the pattern
of occurrence of diseases within the breed
of interest.
There have been few population stud-
ies on canine longevity and causes of
death. Many studies have used veterinary
teaching hospital populations that
invariably include a high proportion of
referral cases. For example, Patronek and
others (1997) reported on mortality data
for 23,535 dogs taken from the Veteri-
nary Medical Data Base in North Amer-
ica. Another study reported age at death
and cause of death taken from necropsy
data from 1962 to 1976 for 2002 dogs
from the Angell Memorial Animal Hos-
pital in Boston (Bronson 1982). A Japa-
nese study reported expected lifespans
from pet cemetery data (Hayashidani and
others 1988). A Swedish group reported
mortality risks and causes of death for
over 200,000 dogs based on records of
a Swedish pet insurance company (Bon-
nett and others 1997; Egenvall and oth-
ers 2000a). This group also reported the
age patterns for risk of death in selected
breeds of dogs insured for life (Egen-
vall and others 2000b). Michell (1999)
reported the results of a questionnaire
survey of UK pet owners in 1999 that
included age at death and cause of death
for 3126 dogs. Longevity estimates for
UK dogs have also been reported using
information from a Pedigree Masterfoods
survey of pet ownership (Reid and Peter-
son 2000). A recent study reported cause
of death and age at death data for 2928
dogs from a survey of members of the
Danish Kennel Club (KC) carried out
in 1997 (Proschowsky and others 2003).
Unfortunately, none of these studies
address longevity and cause of death in
the general canine population and, as
such, have inherent sampling biases.
Results from these studies provide wide-
ranging estimates of lifespan and must be
interpreted with caution.
This survey was initiated follow-
ing meetings with the KC/British Small
Animal Veterinary Association Scien-
tific Committee in order to provide bet-
ter evidence upon which to base advice
to breeders, funding bodies and policy
makers. The overall aims of the survey
were to identify important breed-specif-
ic problems for future genetic research
and to provide baseline information
against which the success of future con-
trol schemes could be measured. The
OBJECTIVES: To collect information on the cause of death and
longevity of dogs owned by members of the numerically largest
breed clubs of 169 UK Kennel Club-recognised breeds.
METHODS: A cross-sectional study was carried out. Approximately
58,363 questionnaires were sent out to breed club members in
2004 (nine clubs failed to report the exact number of question-
naires sent out). Owners reported age at death and cause(s) of
death for all dogs that had died within the previous 10 years.
RESULTS: A total of 13,741 questionnaires (24% response rate)
containing information on 15,881 deaths were included in the
analysis. Breed-specific response rates ranged from 64⋅7 to 4⋅5%.
The median age at death was 11 years and 3 months (minimum=2
months, maximum=23 years and 5 months) and it varied by breed.
The most common causes of death were cancer (n=4282, 27%),
“old age” (n=2830, 18%) and cardiac conditions (n=1770, 11%).
CLINICAL SIGNIFICANCE: This survey shows breed differences in lifespan
and causes of death, and the results support previous evidence that
smaller breeds tend to have longer lifespan compared with larger
breeds. Although many of the breeds in the study may not be rep-
resentative of the general pedigree dog population in the UK, the
results do contribute to the limited information currently available.
htt
p:
//
www.
b
sava.com
/

Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association 513
Results of a health survey of UK purebred dogs
specific aim of this study was to collect
information on the cause of death and
longevity of UK pedigree dogs. Addi-
tional health information, including the
occurrence of disease in 36,006 live dogs,
will be reported separately.
MATERIALS AND METHODS
Sampling frame and data
collection
A cross-sectional study was carried out
to collect health information on pedigree
dogs owned by members of the numeri-
cally largest (“parent”) breed clubs in the
UK. There are over 700 breed clubs in
the UK with a lot of overlap in mem-
bership, particularly among regional
and national clubs. Therefore, to limit
the probability of individual breed club
members receiving multiple question-
naires, only the numerically largest club
for each breed was included in the sur-
vey. The KC also recommended send-
ing out questionnaires to 18 additional
breed clubs for 11 breeds to improve
coverage of those breeds that had a very
small number of clubs and few annual
registrations with the KC or where the
clubs were very large or regionalised. As a
result, one additional club was included
for the beagle, Bernese mountain dog,
border terrier, Finnish spitz, Scottish
terrier and West Highland white terrier
breeds, and two additional clubs were
included for the akita, collie, golden
retriever, Staffordshire bull terrier, Welsh
corgi and Yorkshire terrier breeds.
Packages were sent out to 170 UK KC-
recognised breeds via 188 breed club sec-
retaries. The initial mailing to the breed
club secretaries took place in December
2003 with individual clubs mailing out to
their members starting in January 2004.
The number of questionnaires sent out to
each breed club was based on numbers of
members reported to the KC plus 10%.
Monthly reminders were sent out from
the UK KC via the dog press from Janu-
ary to April with a deadline for return
of completed questionnaires of April 31,
2004. This was extended to June 30 to
allow for slow responders, and completed
questionnaires returned by October 2004
were included.
Owners were sent a postal question-
naire with a self-addressed prepaid return
envelope and they were asked to complete
the questionnaire for all dogs of the breed
for which they belonged to the breed club
that mailed them the questionnaire. The
survey was anonymous in that the ques-
tionnaire did not ask for the owners name
or any other identifying feature.
Questionnaire
The questionnaire consisted of four sec-
tions containing 22 questions in total.
Section A asked about all dogs that were
alive and currently owned by these own-
ers and contained 14 questions cover-
ing 12 body systems, cancer and other
conditions. Section B contained seven
questions on breeding of females that
had been owned and bred in the previ-
ous 10 years and included questions on
the numbers of litters and puppies born.
Section C contained one question on the
occurrence of birth defects and conditions
affecting puppies in the first 8 weeks of
life during the previous 10 years. Section
D contained one question on the age at
death and cause(s) of death of all dogs that
had died within the previous 10 years.
Questions were developed, pretested and
refined for use in a pilot study of two
breeds (Norfolk terrier and German spitz).
Owners were asked to be as specific as pos-
sible when describing a disease condition,
cause of death or congenital condition,
using the diagnosis made by their veteri-
nary surgeon whenever possible. We also
suggested that owners consider contacting
their veterinary surgeon to ask for help if
they had difficulty remembering the diag-
nosis. Veterinary surgeons in the UK were
informed of the survey via a letter to The
Veterinary Record at the start of the study.
An appendix list of names and synonyms
of example conditions was included for
each body system question in section A as
well as for congenital conditions in section
C. The questionnaire, appendix, over-
all results and breed-specific reports are
available on the UK KC website at http://
www.the-kennel-club.org.uk/.
Data processing
Specialised information capture software
(Verity [was Cardiff ] TELEform®) was
used to scan the returned questionnaires.
Data entry involved scanning and verify-
ing of the electronically acquired data. The
scanned and verified data were exported
into an Access® (Microsoft) database for
checking and recoding. When the verifier
module of Teleform checks and codes the
scanned data, the software is set so that if
the system reads a number with a certain
accuracy, it will not ask for the informa-
tion to be verified. For example, with age
at death, when the software was 90% cer-
tain that what it had scanned in was the
number 4, then it recorded the number
scanned as a 4 and did not bring the num-
ber up for verification. For the numerical
information, during the initial scanning of
the forms, almost 100 forms were scanned
and verified several times, allowing us to
look at the reliability of the scanning soft-
ware. Although the exact error rates were
not recorded, the software appeared to be
very reliable in that it recorded the same
information each time a form was repeat-
edly scanned. For the free text boxes, what
was written in the box was entered manu-
ally during the verification process as the
system did not reliably scan handwritten
text. The coding of the causes of death
was done by choosing a cause of death
from a drop down list developed by the
first author before data entry began. As
data entry progressed and the first forms
were repeatedly scanned, the list of pos-
sible causes of death was lengthened and
refined as necessary by the first author.
When it was not clear what the cause of
death should be coded as, the form was
marked “for review” and the first author
reviewed and coded these causes of death
and added any required new terms to the
list used for coding. This was an itera-
tive process that continued through the
first several days of data entry and then
as needed throughout the remainder of
data entry. A similar system of using lists
of diagnoses for the various body systems
was used for the disease conditions affect-
ing live dogs.
Diagnostic categories were devel-
oped for cause of death by firstly group-
ing diseases by organ system affected.
Organ system categories included car-
diac, cardiovascular, cerebral vascular,
dermatologic, endocrine, gastrointestinal,
hepatic, musculoskeletal, neurologic, ocu-
lar, reproductive, respiratory and urologic

V. J. Adams and others
514 Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association
as used on the questionnaire. When a
cause of death was interpreted to affect
more than one body system and a diag-
nosis could be attributed to such a cause,
separate categories were included for these
specific causes of death. Thus, additional
categories were created for behaviour,
cancer, collapse, drowning, hyperther-
mia, immune-mediated, infection, inter-
nal haemorrhage, perioperative, pining,
poisoning, portosystemic shunt, senility,
septicaemia, sudden death, systemic and
trauma. A category for combinations was
used when two or more specifically code-
able causes of death were stated for one
dog. A category for “old age” was used
when either age or “old age” was stated
as the sole cause of death or when age
was stated together with another condi-
tion such as heart failure, kidney failure,
arthritis, incontinence, colitis or senility.
Dogs with senility or dementia stated
as the sole cause of death were coded as
senility. A category called unknown was
created when the stated cause of death
could not be categorised into any one of
the above groups (“uncodeable” causes of
death) or when the words “unknown,”
“undiagnosed,” “died” or “sudden death”
were used. Uncodeable causes of death
included those conditions reported by
owners that could not be placed into any
of the other categories, such as weakness,
loss of limb function or weight loss.
Statistical analysis
Longevity results, as overall and breed-
specific age at death, are reported as
median (range), as is appropriate for
skewed data. Age at death for groups of
dogs was compared using non- parametric
tests (Wilcoxon rank sum for two
groups). The association between age
at death and the size of the breed was
assessed using published values for aver-
age “ideal” bodyweight (Alderton 1993).
Linear regression was used to examine the
association between median age at death
and ideal bodyweight with the hypoth-
esis being that larger dog breeds would
have a shorter average lifespan compared
with smaller dog breeds. Descriptive sta-
tistics and hypothesis testing were per-
formed using SPSS (Statistical Package
for the Social Sciences v. 13.0, SPSS
Inc. 2004, Chicago, IL, USA). The level
of significance was set at P<0⋅05 for all
hypothesis tests.
The most common causes of death
are reported overall and as breed-specific
proportional mortalities for the four most
commonly affected organ systems/catego-
ries. Breed-specific proportional mortali-
ties were calculated as the number of deaths
affecting a specific organ system divided by
the total number of deaths for a specific
breed and are reported with 95% confi-
dence limits. Breeds with a response rate of
>15% and having had >50 deaths reported
were included as separate breeds, and the
remaining breeds were combined into
“other breeds” in order to avoid reporting
less accurate results for breeds with very
low response rates and/or very few deaths
reported. Proportional mortality computa-
tions were performed using SAS (version
8, SAS Institute 2000, Cary, NC, USA).
RESULTS
Response rate
A total of 72,832 questionnaires were
sent out to breed club secretaries, and
approximately 58,363 questionnaires
were sent out to their members. As nine
clubs failed to report the exact number
of questionnaires sent out to their mem-
bers, the exact number of questionnaires
sent out was unknown. Approximately
600 (4%) questionnaires had data that
were unuseable and not entered (due to
failure to report the breed, reporting on
>1 breed on a questionnaire or unsolicited
photocopied questionnaires). A total of
13,791 questionnaires were entered (24%
response rate), of which 13,759 were use-
able (32 questionnaires were entered in
duplicate and were removed from the
database). The responses for the German
shepherd dog were excluded from further
analyses as 1425 questionnaires were sent
out to the British Association for German
shepherd dogs, but only 18 question-
naires were returned and the breed club
did not report how many questionnaires
were sent out. After excluding German
shepherd dogs, 13,741 questionnaires
were available for further analysis. Breed-
specific response rates ranged from a high
of 64⋅7% to a low of 4⋅5% with a median
of 23⋅9% (Table 1).
There were 13,367 questionnaires
with information on 36,006 live dogs,
7125 of which also reported deaths, and
there were 374 questionnaires that only
reported deaths to give a total of 7499
forms with information on 15,881 deaths.
Deaths were reported for 165 breeds with
the Bergamasco, Bolognese, Ibizan hound
and Lagotto Romagnolo each having no
reported deaths. There were 72 breeds
with a response rate of >15% and with >50
deaths reported, and there were 93 breeds
included in the “other breeds” category
for estimation of proportional mortalities.
Overall, 931 of the 15,881 reported deaths
(5⋅9%) had a post-mortem examination.
For those dogs that underwent a post-
mortem examination, the median age at
death was 7⋅17 years (00·17 to 20·00), and
20% (191) were made up of five breeds:
flatcoated retriever (50), golden retriever
(42), Labrador retriever (35), cavalier King
Charles spaniel (34) and Bernese moun-
tain dog (30). The median age at death for
those dogs that underwent a post-mortem
examination was significantly lower than
the median age at death for those that
did not (11⋅5 years, P<0⋅0001). The most
commonly reported cause of death in
those dogs that underwent a post-mortem
was cancer (337), although the majority of
these were still reported as type unspeci-
fied. Only 17 of the 597 dogs that were
reported to have died due to a stroke, and
19 of 291 dogs that were reported to have
died of a heart attack, underwent post-
mortem examination and it was not clear
from what was reported whether the post-
mortem confirmed the diagnosis. Owners
did not indicate whether the post-mortem
helped to provide the reported cause of
death or not.
Lifespan
The overall median age at death was
11 years and 3 months (2 months to 23
years and 5 months; Fig 1). Only 20% of
dogs remained alive at 14 years of age and
this had fallen to <10% by 15 years. Breed-
specific age at death is shown in Table 1.
Median age at death was significantly
negatively correlated with average ideal
bodyweight with 40% of the variability in
age at death explained by bodyweight for
81 breeds with published values for ideal
bodyweights (P<0⋅0001; Fig 2).

Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association 515
Results of a health survey of UK purebred dogs
Table 1. For each of the 165 breeds experiencing a total of 15,881 deaths over the 10 years before 2004: the number of
questionnaires sent out and returned for each breed, br eed-specific questionnaire response rates, number and percent of
deaths, and minimum, median and maximum age at death as well as the Kennel Club breed group
Breed name Questionnaires Deaths Age at death
Breed weight
group†
Numbers
sent
Number
of returns % RR*
Number
of deaths
% of all
deaths Minimum Median Maximum
Affenpinscher 225 66 29⋅3210⋅10 0⋅25 11⋅42 15⋅17 2
Afghan hound 206 58 28⋅2 143 0⋅90 0⋅83 11⋅92 16⋅58 4
Airedale terrier 260 66 25⋅4810⋅50 0⋅67 10⋅75 16⋅17 3
Akita 176 23 13⋅1280⋅20 0⋅33 9⋅92 13⋅67 4
Alaskan Malamute 185 57 30⋅8140⋅10 2⋅17 10⋅71 13⋅53
American cocker spaniel 198 33 16⋅7600⋅40 0⋅17 10⋅33 17⋅33 3
Anatolian/Karabash 180 22 12⋅2230⋅10 4⋅42 10⋅75 13⋅42 4
Australian cattle dog 83 22 26⋅5110⋅10 1⋅511⋅67 15⋅92 3
Australian shepherd 132 49 37⋅1220⋅10 3 9 15 3
Australian silky terrier 37 7 18⋅950⋅00 11⋅08 14⋅25 15⋅33 1
Australian terrier 74 12 16⋅2110⋅10 3⋅92 12⋅08 15 1
Basenji 171 40 23⋅4460⋅30 0⋅92 13⋅54 17⋅52
Basset Fauve de Bretagne 125 41 32⋅8150⋅10 0⋅92 10⋅42 13⋅92 3
Basset Griffon Vendeen 306 135 44⋅1760⋅50 1⋅25 12⋅04 17⋅33 3
Basset hound 500 88 17⋅6 142 0⋅90 0⋅25 11⋅29 16⋅67 3
Beagle 585 177 30⋅3 241 1⋅50 1⋅17 12⋅67 17⋅75 2
Bearded collie 729 239 32⋅8 278 1⋅80 0⋅33 13⋅519⋅53
Bedlington terrier 200 61 30⋅5480⋅30 1⋅17 13⋅38 18⋅42 2
Belgian shepherd 339 98 28⋅9 113 0⋅70 1⋅512⋅518⋅17 4
Bergamasco‡ 14 4 28⋅6‡ ‡ 4
Bernese mountain dog 1200 361 30⋅1 394 2⋅50 0⋅42 8 15⋅17 4
Bichon frise 161 34 21⋅1340⋅20 3⋅25 12⋅92 16⋅42 1
Bloodhound 180 46 25⋅6820⋅50 0⋅92 6⋅79 12⋅08 5
Bolognese‡ 36 2 5⋅6‡ ‡ 3
Border collie 558 96 17⋅2 106 0⋅70 0⋅17 12⋅25 17⋅33 3
Border terrier 528 152 28⋅8 177 1⋅10 0⋅17 14 22⋅08 2
Borzoi 182 42 23⋅1870⋅50 0⋅58 9⋅08 14⋅25 4
Boston terrier 110 25 22⋅7420⋅30 5⋅17 10⋅92 15⋅75 2
Bouvier Des Flandres 140 45 32⋅1390⋅20 0⋅75 11⋅33 18⋅33 4
Boxer 450 68 15⋅1 130 0⋅80 0⋅33 10⋅25 15⋅25 4
Bracco Italiano§ 43 11 25⋅610⋅00 2⋅67§3
Briard 238 75 31⋅5710⋅40 1 11⋅17 16⋅75 4
Brittany 225 59 26⋅2280⋅20 0⋅67 12⋅88 16⋅08 3
Bull terrier 864 156 18⋅1 209 1⋅30 0⋅51018⋅53
Bulldog 841 143 17⋅0 180 1⋅10 0⋅83 6⋅29 14⋅42 3
Bullmastiff 491 80 16⋅3960⋅60 1⋅33 7⋅46 13⋅54
Cairn terrier 397 96 24⋅2 124 0⋅80 0⋅25 14 18⋅33 2
Canaan dog 35 15 42⋅920⋅00 13⋅08 14⋅63 16⋅17 3
Cavalier King Charles
spaniel 1150 306 26⋅6 682 4⋅30 0⋅17 11⋅38 18⋅52
Cesky terrier 65 13 20⋅090⋅10 5⋅58 8⋅42 12⋅67 2
Chesapeake Bay retriever 277 58 20⋅9450⋅30 0⋅75 10⋅75 15⋅67 3
Chihuahua 407 37 9⋅1710⋅40 0⋅17 12⋅42 19⋅83 1
Chinese crested 68 13 19⋅1140⋅10 4⋅83 10⋅08 16⋅17 1
Chow Chow 345 23 6⋅7800⋅50 0⋅33 9⋅38 14⋅17 3
Clumber spaniel 388 54 13⋅9690⋅40 0⋅92 10⋅33 16 4
(continued overleaf)

V. J. Adams and others
516 Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association
Cocker spaniel 1000 206 20⋅6 289 1⋅80 0⋅42 11⋅17 17⋅25 3
Collie 329 42 12⋅8790⋅50 0⋅17 12⋅67 17⋅08 3
Curly coated retriever 100 35 35⋅0400⋅30 0⋅92 10⋅75 15 4
Dachshund 810 155 19⋅1 245 1⋅50 0⋅33 12⋅67 19 2
Dalmatian 686 210 30⋅6 199 1⋅30 0⋅17 12⋅5174
Dandie Dinmont terrier 207 65 31⋅4620⋅40 4 12⋅17 17⋅75 2
Deerhound 570 238 41⋅8 287 1⋅80 0⋅17 8⋅67 16⋅75 5
Dobermann 300 58 19⋅3 100 0⋅60 1⋅08 10⋅516⋅54
Dogue de Bordeaux 249 41 16⋅550⋅00 0⋅67 3⋅83 16 4
English setter 952 254 26⋅7 384 2⋅40 0⋅83 11⋅58 20 4
English springer spaniel 353 71 20⋅1900⋅60 1⋅17 12 19⋅53
English toy terrier 152 28 18⋅4190⋅10 1⋅42 12 15⋅92 1
Estrela mountain 35 14 40⋅030⋅00 9⋅75 9⋅75 12⋅83 3
Field spaniel 235 56 23⋅8680⋅40 1⋅25 11⋅63 17⋅25 3
Finnish Lapphund 180 28 15⋅650⋅00 0⋅67 7⋅33 10⋅25 2
Finnish spitz 133 51 38⋅3420⋅30 1⋅511⋅13 15⋅92 3
Flat-coated retriever 1556 472 30⋅3 610 3⋅80 0⋅33 9⋅83 16⋅92 4
Fox terrier 239 45 18⋅8440⋅30 0⋅83 13⋅13 17⋅25 2
French bulldog 334 72 21⋅6710⋅40 0⋅42 9 14⋅67 2
German longhaired
pointer 42 12 28⋅620⋅00 8 10⋅5133
German pinscher 55 24 43⋅6240
⋅20 5⋅92 11⋅38 17⋅92 2
German shor thaired
pointer 680 192 28⋅2 159 1⋅00 0⋅92 12 17 4
German spitz 165 55 33⋅3430⋅30 1⋅08 11⋅33 16⋅33 2
German wirehaired pointer 58 66 113⋅8410⋅30 0⋅92 10 15⋅58 4
Giant schnauzer 190 69 36⋅3390⋅20 0⋅67 10 16⋅92 4
Glen of Imaal terrier 93 35 37⋅660⋅00 7⋅75 10⋅42 14⋅53
Golden retriever 3282 538 16⋅4 927 5⋅80 0⋅42 12⋅25 17⋅25 4
Gordon setter 545 126 23⋅1 157 1⋅00 0⋅33 11⋅08 16⋅25 3
Grand Bleu de Gascoigne 21 7 33⋅360⋅00 2⋅54⋅54 10⋅08 3
Great Dane 360 72 20⋅0 171 1⋅10 0⋅33 6⋅514⋅17 5
Greenland dog 11 3 27⋅320⋅00 7⋅58⋅46 9⋅42 3
Greyhound 75 33 44⋅0690⋅40 0⋅33 9⋅08 14⋅17 4
Griffon Bruxellois 207 72 34⋅8710⋅40 0⋅42 12 19⋅75 1
Hamiltonstovare 60 15 25⋅060⋅00 3⋅67 10⋅13 16⋅08 3
Havanese 50 11 22⋅030⋅00 4 10⋅25 18⋅17 3
Hovawart 110 33 30⋅0190⋅10 7 12⋅92 14⋅75 3
Hungarian Puli 98 30 30⋅6260⋅20 1 12⋅42 17⋅92 3
Hungarian vizsla 300 36 12⋅0380⋅20 0⋅512⋅92 17 4
Hungarian wirehaired
vizsla 136 54 39⋅7150⋅10 2 9⋅83 15 3
Ibizan hound‡ 44 2 4⋅5‡ ‡ 2
Irish red & white setter 215 139 64⋅7 179 1⋅10 0⋅92 11⋅42 17⋅53
Irish setter 1298 235 18⋅1 451 2⋅80 0⋅33 12 17⋅08 4
Irish terrier 150 11 7⋅320⋅00 14⋅514⋅83 15⋅17 2
Irish water spaniel 210 90 42⋅9950⋅60 0⋅58 9⋅33 17⋅17 4
Irish wolfhound 261 40 15⋅3 112 0⋅70 1⋅17 7⋅04 11⋅83 5
Italian greyhound 287 42 14⋅6460⋅30 0⋅513⋅5181
Table 1. (Continued)
Breed name Questionnaires Deaths Age at death
Breed weight
group†
Numbers
sent
Number
of returns % RR*
Number
of deaths
% of all
deaths Minimum Median Maximum
(continued overleaf)

Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association 517
Results of a health survey of UK purebred dogs
Italian Spinone 376 122 32⋅4470⋅30 0⋅58 9 16 4
Japanese Chin 158 25 15⋅8380⋅20 0⋅58 9⋅25 14⋅08 1
Japanese spitz 100 8 8⋅0100⋅10 1 12⋅29 15 1
Keeshond 275 78 28⋅4 104 0⋅70 1⋅83 12⋅21 16⋅08 3
Kerr y Blue terrier 120 16 13⋅3200⋅10 3 11⋅515⋅17 3
King Charles spaniel 170 24 14⋅1500⋅30 1 10⋅04 23⋅42 2
Komondor 27 10 37⋅0100⋅10 3 9⋅13 13⋅17 5
Kooikerhondje 35 9 25⋅770⋅00 0⋅53⋅92 13⋅92 3
Labrador retriever 1775 369 20⋅8 574 3⋅60 0⋅17 12⋅25 19 4
Lagotto Romagnolo‡ 205 33 16⋅1‡ ‡ 3
Lakeland terrier 183 26 14⋅2140⋅10 2⋅08 15⋅46 18⋅08 2
Lancashire heeler 208 66 31⋅7300⋅20 0⋅17 11⋅75 21⋅83 2
Large Munsterlander 220 99 45⋅0690⋅40 1⋅511⋅33 16⋅58 4
Leonberger 350 109 31⋅1470⋅30 0⋅42 7⋅08 12⋅67 5
Lhasa apso 225 52 23⋅1840⋅50 1 14⋅33 18⋅42 2
Lowchen 75 12 16⋅090⋅10 1⋅08 10 16⋅51
Maltese 247 37 15⋅0420⋅30 1⋅08 12⋅25 19⋅17 1
Manchester terrier 178 59 33⋅1320⋅20 0⋅92 12⋅83 17⋅51
Maremma sheepdog 64 13 20⋅3190⋅10 1⋅17 10 14⋅54
Mastiff 424 51 12⋅0800⋅50 0⋅75 6⋅83 16⋅33 5
Miniature bull terrier 95 25 26⋅3200⋅10 0⋅67 6⋅08 13⋅17 2
Miniature pinscher 249 30 12⋅0270⋅20 2⋅51317⋅92 1
Miniature poodle 81 26 32⋅1230⋅10 6⋅513⋅92 18⋅52
Miniature schnauzer 583 289 49⋅6 214 1⋅30 0⋅58 12⋅08 18⋅17 2
Neopolitan mastiff 80 9 11⋅370⋅00 0⋅67 2⋅33 16⋅08 4
Newfoundland 890 233 26⋅2 269 1⋅70 0⋅42 9⋅67 15⋅83 5
Norfolk ter rier 598 233 39⋅0 189 1⋅20 1⋅511172
Norwegian buhund 84 29 34⋅5170⋅10 3⋅512⋅67 16⋅17 2
Norwegian elkhound 268 71 26⋅5710⋅40 0⋅25 13⋅17 17⋅17 3
Norwich ter rier 200 71 35⋅5560⋅40 1⋅33 13⋅38 16⋅75 2
Nova Scotia duck tolling
retriever 175 60 34⋅390⋅10 0⋅75 8 14⋅33 3
Old English sheepdog 66 32 48⋅5650⋅40 1⋅33 10⋅75 15⋅08 4
Otterhound 139 20 14⋅4540⋅30 1 10⋅21 15 4
Papillon/butterfly dog 475 78 16⋅4570⋅40 1⋅25 13⋅08 19 1
Parson Russell terrier 248 37 14⋅9170⋅10 0⋅25 13 17⋅17 2
Pekingese 174 33 19⋅0870⋅50 1 11⋅42 18⋅17 1
Pharoah hound 30 8 26⋅7150⋅10 3 11⋅83 17⋅17 3
Pointer 386 93 24⋅1 145 0⋅90 0⋅512⋅42 16⋅42 3
Polish lowland sheepdog 134 21 15⋅7110⋅10 2⋅83 9⋅58 15⋅08 3
Pomeranian 141 22 15⋅6290⋅20 0⋅58 9⋅67 17⋅25 1
Portuguese water dog 46 11 23⋅960⋅00 4⋅25 11⋅42 15⋅25 3
Pug 1200 196 16⋅3 163 1⋅00 0⋅17 11 17 2
Pyrenean mountain dog 539 74 13⋅7660⋅40 0⋅75 9⋅58 16 5
Pyrenean sheepdog 22 3 13⋅640⋅00 3⋅67 5⋅79 7 4
Rhodesian ridgeback 535 172 32⋅1 183 1⋅20 1 11 16⋅17 4
Rottweiler 252 67 26⋅6 137 0⋅90 0⋅42 8⋅92 16⋅75 4
Russian black terrier 111 23 20⋅740⋅00 0⋅25 1⋅79 11⋅53
Table 1. (Continued)
Breed name Questionnaires Deaths Age at death
Breed weight
group†
Numbers
sent
Number
of returns % RR*
Number
of deaths
% of all
deaths Minimum Median Maximum
(continued overleaf)

V. J. Adams and others
518 Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association
Saluki/gazelle hound 232 74 31⋅9 132 0⋅80 1⋅92 12 16⋅33 4
Samoyed 425 173 40⋅7 223 1⋅40 0⋅33 12⋅5174
Schipperke 109 21 19⋅3360⋅20 0⋅51317⋅52
Schnauzer (standard) 860 76 8⋅8520⋅30 1 11⋅96 20⋅08 3
Scottish terrier 278 47 16⋅9590⋅40 0⋅510⋅25 17⋅52
Sealyham terrier 100 21 21⋅0120⋅10 2⋅17 12⋅25 14⋅92 2
Shar pei 447 47 10⋅5600⋅40 1⋅92 6⋅29 16⋅53
Shetland sheepdog 1003 207 20⋅6 364 2⋅30 0⋅17 12⋅5192
Shiba Inu (Japanese) 78 17 21⋅830⋅00 4⋅5792
Shih-tzu 265 63 23⋅8830⋅50 1⋅513⋅17 19⋅33 2
Siberian husky 956 188 19⋅7 129 0⋅80 0⋅42 12⋅58 18⋅08 3
Skye terrier 130 28 21⋅5370⋅20 1⋅92 11 15⋅58 2
Sloughi§ 30 8 26⋅710⋅00 5⋅00§ 4
Soft coated wheaten
terrier 824 380 46⋅1 111 0⋅70 0⋅512⋅5173
St Bernard 130 24 18⋅5530⋅30 0⋅5712⋅75 5
Staffordshire bull terrier 833 132 15⋅8 117 0⋅70 1⋅17 12⋅75 17⋅17 3
Standard poodle 237 83 35⋅0 118 0⋅70 1⋅25 12 18 4
Sussex spaniel 191 60 31⋅4420⋅30 1⋅08 11⋅13 16⋅53
Swedish vallhund 71 16 22⋅5170⋅10 4 14⋅42 18⋅83 2
Tibetan mastiff 60 23 38⋅3100⋅10 3⋅58 11⋅92 17⋅17 4
Tibetan spaniel 314 99 31⋅5 125 0⋅80 1⋅514⋅42 19⋅08 2
Tibetan terrier 314 103 32⋅8950⋅60 0⋅512⋅17 18⋅25 2
Toy poodle 48 11 22⋅9200⋅10 1⋅75 14⋅63 18⋅92 1
Weimaraner 1296 357 27⋅5 242 1⋅50 0⋅67 11⋅13 18⋅83 4
Welsh corgi Cardigan∞53 0⋅30 4⋅08 12⋅17 16⋅52
Welsh corgi Pembroke∞416 95 22⋅8 116 0⋅70 1⋅42 12⋅21 17⋅58 2
Welsh springer spaniel 560 190 33⋅9 157 1⋅00 1 12⋅58 16⋅67 3
Welsh terrier 238 66 27⋅7230⋅10 0⋅512⋅67 18 2
West Highland white
terrier 628 89 14⋅2 127 0⋅80 0⋅58 13 18⋅17 2
Whippet 846 374 44⋅2 486 3⋅10 0⋅17 12⋅79 18⋅17 2
Yorkshire ter rier 276 26 9⋅4460⋅30 0⋅25 12⋅67 17⋅33 1
Total 15,881 100⋅00 0⋅17 11⋅25 23⋅42
*RR = response rate = (number of returned forms/number of forms sent out)×100
†1 = toy; 2 = small; 3 = medium; 4 = large; 5 = giant for breed weight group
‡Bergamasco, Bolognese, Ibizan hound and Lagotto Romagnolo each had no repor ted deaths
§Bracco Italiano and Sloughi each had only one reported death
∞Questionnaires were sent out to the Welsh Corgi Club and League but returns were separated into Welsh Corgi Cardigan and Welsh Cor gi Pembroke for further analysis
Table 1. (Continued)
Breed name Questionnaires Deaths Age at death
Breed weight
group†
Numbers
sent
Number
of returns % RR*
Number
of deaths
% of all
deaths Minimum Median Maximum
Of the 14 breeds with the highest medi-
an age at death (≥13⋅5 years), 21% were
toy, 64% were small and 14% medium
(Table 2). Long-lived breeds were reported
to die of diseases normally associated with
ageing, including cancer and chronic renal
failure. Of the 11 breeds with the lowest
median age at death (<8 years), it is not
unexpected that 6 (55%) were giant and
2 (18%) were large (Table 3). Addition-
ally, 2 (18%) were medium breeds, the
British bulldog and the shar pei, both
with a median age at death of only 6⋅3
years. Short-lived breeds succumbed more
frequently to cardiac diseases, particularly
cardiomyopathy and valvular disease, and
gastrointestinal diseases, particularly gas-
trointestinal dilatation/volvulus.
Causes of death
The three most commonly reported
causes of death overall were cancer
(n=4282, 27%), “old age” (n=2829, 18%)
and cardiac conditions (n=1770, 11%).
The most common causes of death overall
are presented in Table 4. A category called
“other” was used for the remaining causes
of death that affected 3% of dogs and
included, in descending order, senility,
internal bleeding, collapse, infection, ocu-
lar, dermatologic, portosystemic shunt,
septicaemia, systemic, cardiovascular, pin-
ing, hyperthermia and drowning. Greater

Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association 519
Results of a health survey of UK purebred dogs
242220181614121086420
Age at death in years
2,000
1,800
1,600
1,400
1,200
1,000
800
600
400
200
0
Number of dogs
FIG 1. Histogram of the overall longevity for 165 breeds with reported deaths: overall median age
at death was 11 years and 3 months (minimum=2 months, maximum=23 years and 5 months);
mean age at death was 10 years and 6 months (sd=4 years)
706050403020100
Avera
g
e published ideal bod
y
wei
g
ht (kilo
g
rams)
16
14
12
10
8
6
4
2
0
Median age at death (years)
FIG 2. Association of median age at death with average ideal bodyweight by breed (age at death in
years = 12⋅6−0⋅08×bodyweight in kilograms, r=0⋅4, P<0⋅0001)
than 75% of the cancers were reported as
having an unspecified type (3319 of 4282;
Table 4). For these tumours of unspecified
type, the most common locations were
the liver (339), mammary glands (278)
and brain (234) with the location being
unspecified in 545 dogs. When the type
of tumour was specified, lymphosarcoma
was the most frequently reported type
(368 dogs), followed by carcinoma (128
dogs). Owners did not report whether
they had contacted their veterinary sur-
geon to ask for help if they had difficulty
remembering a specific diagnosis of cause
of death.
Breed-specific proportional mortali-
ties are presented for the four most com-
mon causes of death in 72 breeds (Table
5). The breeds with the highest propor-
tional mortalities for cancer included, in
descending order, Irish water spaniel, Flat-
coated retriever, Hungarian wirehaired
vizsla, Bernese mountain dog, Rottweiler,
Italian Spinone, Leonberger, Stafford-
shire bull terrier, Welsh terrier and giant
schnauzer. Breeds with “old age” as the
highest breed-specific proportional mor-
tality included, in descending order, Lhasa
apso, Manchester terrier, border terrier,
Norwich terrier, cairn terrier, papillon,
Tibetan spaniel, Dalmatian, whippet and
bearded collie. Breeds with a cardiac con-
dition as the highest breed-specific pro-
portional mortality, in descending order,
included cavalier King Charles spaniel,
Norfolk terrier, deerhound, Griffon Brux-
ellois and British bulldog.
Discussion
The median age at death in the pres-
ent study of 11⋅25 years is similar to the
median age at death of 12 years found
in an earlier UK study (Michell 1999)
and 10 years found in a Danish KC sur-
vey (Proschowsky and others 2003). All
of these median ages at death are higher
than the mean of 6⋅4 years found in
USA necropsy study (Bronson 1982),
and this is likely due to the much differ-
ent population of dogs that would have
been presented to a Veterinary Teaching
Hospital for post-mortem examination.
Such a population would be likely to have
included more sudden deaths and deaths
due to trauma which often occur at a
younger age.

V. J. Adams and others
520 Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association
Similar to other studies, smaller breeds
such as the terriers, beagle, bearded collie,
bichon frise, dachshund, miniature pin-
scher, miniature poodle, toy poodle and
whippet lived longer, and giant breeds
such as Bernese mountain dog, deer-
hound, Leonberger, Irish wolfhound,
St Bernard, mastiff and bloodhound had
shorter lifespan (Bronson 1982; Michell
1999; Proschowsky and others 2003).
These results and the negative correlation
of age at death with bodyweight were as
expected based on previous studies. The
ideal bodyweight was used here as there
are no published studies that report aver-
age bodyweights for the breeds included in
this study. The same results were obtained
when the breeds were categorised accord-
ing to body size into giant, large, medium,
small and toy breeds (breed weight group
in Table 1). Interestingly, three medium-
sized breeds had the shortest lifespan with
a median age at death of 6⋅3 years for the
british bulldog (180 deaths) and shar pei
(60 deaths) and 6⋅08 years for the minia-
ture bull terrier (20 deaths).
Due to the difficulties associated
with random sampling of the pedigree
dog population in the UK, this study
took the approach of obtaining a conve-
nience sample. Thus, the results of this
study cannot necessarily be generalised
to all dogs or breeds. The breed-specific
response rates were highly variable and
affect the generalisability of the results
on a breed by breed basis. The limita-
tions of owner-reported causes of death
means that some of the results must be
interpreted with caution. For example,
reported causes of death due to a stroke
or cerebral vascular accident were not
supported by a post-mortem examination
in 97⋅5% of cases in this study. Strokes are
relatively uncommon with less than 50
cases of brain infarction in dogs reported
in the veterinary literature to date (Garosi
and others 2005). The prognosis for brain
infarction is considered to be fair to good
as more than 66% of the 33 dogs reported
in the study by Garosi and others (2005)
either returned to normal or experienced
considerable improvement in their neu-
rologic status. Although it appears that
a stroke rarely kills, a suspected or con-
firmed stroke could leave a dog in a con-
dition where euthanasia was chosen as the
most humane method of treatment. The
same could also be said for the reporting
of heart attack as an unconfirmed cause of
death in this study. Since only a small pro-
portion of the dogs that were reported to
have died due to a stroke or heart attack
underwent post-mortem examination and
it was not clear whether the post-mortem
confirmed the cause of death, these diag-
noses must be considered with caution.
Given that the most commonly reported
cause of death overall and in those dogs
that underwent a post-mortem was cancer
and that the majority of these were still
reported with an unspecified cell type,
this suggests that either tissues were not
sent to a laboratory for histopathologi-
cal diagnosis or that the owners were not
told of the final diagnosis. An additional
limitation is the bias associated with ask-
ing owners about dogs that had died in
the previous 10 years (recall bias). We
attempted to minimise the effect of recall
bias by asking owners to be as specific as
possible and to contact their veterinary
surgeon if they had difficulty remember-
ing the cause of death. The final limita-
tion is that the survey was anonymous
in order to encourage participation and
this meant that we were not able to verify
cause of death information with the own-
ers or their veterinary surgeons. The three
most commonly reported causes of death
in the present study, cancer, “old age” and
cardiac conditions, were slightly different
from the top three causes of death in the
Danish KC study, age (n=609, 20⋅8%),
cancer (n=425, 14⋅5%) and behaviour
(n=188, 6⋅4%) based on 2650 purebred
and 278 mixed breed dogs (Proschowsky
Table 2. The 14 breeds with the highest median age at death (>
−13⋅5 years)
showing the Kennel Club breed group that these breeds belong to, the
published values for average “ideal” bodyweight (Alderton, 1993), number of
deaths in each breed and median age at death
Breed group Breed Average
bodyweight (kg)
Median age at
death (year)
Number of
deaths
Small Lakeland terrier 5⋅515⋅46 14
Small Irish terrier 11⋅514⋅83 2
Medium Canaan dog 20⋅514⋅63 2
Toy Toy poodle 5⋅014⋅63 20
Small Swedish vallhund 13⋅014⋅42 17
Small Tibetan spaniel 5⋅514⋅42 125
Small Lhasa apso 6⋅514⋅33 84
Toy Australian silky terrier 4⋅514⋅25 5
Small Border terrier 6⋅014⋅00 177
Small Cairn terrier 7⋅014⋅00 124
Small Miniature poodle 13⋅013⋅92 23
Small Basenji 10⋅013⋅54 46
Medium Bearded collie 22⋅513⋅50 278
Toy Italian greyhound 4⋅013⋅50 46
Table 3. The 11 breeds with the lowest median age at death (<8 years) showing
the Kennel Club breed group that these breeds belong to, the published values
for average “ideal” bodyweight (Alderton, 1993), number of deaths in each
breed and median age at death
Breed group Breed Average
bodyweight (kg)
Median age at
death (year)
Number of deaths
Large Bullmastiff 50⋅07⋅46 96
Giant Leonberger 42⋅07⋅08 47
Giant Irish wolfhound 47⋅57⋅04 112
Giant St Bernard 70⋅07⋅00 53
Small Shiba Inu (Japanese) 11⋅57⋅00 3
Giant Mastiff 83⋅06⋅83 80
Giant Bloodhound 43⋅06⋅79 82
Giant Great Dane 50⋅06⋅50 171
Medium Bulldog 24⋅06⋅29 180
Medium Shar pei 18⋅06⋅29 60
Large Dogue de Bordeaux 47⋅53⋅83 5

Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association 521
Results of a health survey of UK purebred dogs
Table 4. The most frequently reported causes of death overall for 165 breeds experiencing 15,881 deaths
Cause of death N % Specific causes reported
Cancer*4282 27⋅0Unspecified type (3319), lymphosarcoma (368), carcinoma (128), other (467)
“Old age” 2829 17⋅8Either “age” or “old age” or one of these in combination with another cause
Cardiac† 1770 11⋅1Failure (815), unspecified defect (310), attack (291), CMP (136), MVD (102), cardiomegaly (78), congenital
defect (18), murmur (11), other (9)
Unknown 828 5⋅2“Unknown,” “undiagnosed,” “died,” “sudden death” (71), other uncodeable causes of death
Urologic 783 4⋅9Unspecified urinary tract disease (468), chronic or unspecified renal failure (232), acute renal failure (21),
other (62)
Combinations 723 4⋅6 Combinations including kidney (181) ± cardiac (164) ± liver (151) most commonly
Neurologic 673 4⋅2Seizure disorder (328) – [“epilepsy” (174), unspecified seizures (154)], interver tebral disc disease (62),
Wobblers (15), unspecified (129), other (139)
Gastrointestinal‡ 652 4⋅1Bloat/GDV (388), pancreatitis (53), unspecified (35), foreign body (24), colitis/diarrhoea (23), HGE (21),
unspecified GE (20), megaoesophagus (16), other (72)
Cerebrovascular 615 3⋅9Stroke or cerebral vascular accident or CVA
Trauma 385 2⋅4Road traffic accident (195), unspecified (133), spinal trauma (29), attacked by dog (17)
Hepatic 336 2⋅1Chronic or unspecified failure (232), unspecified disease (71), acute failur e (21), other (12)
Musculoskeletal 272 1⋅7Ar thritis/osteoarthritis/DJD (167), hip dysplasia (54), other (51)
Endocrine 234 1⋅5Hyperadrenocor ticism (106), diabetes mellitus (76), hypoadrenocor ticism (38), other (14)
Perioperative 222 1⋅4Perioperative or anaesthetic-related (121), postoperative (101)
Behaviour 209 1⋅3Aggression (151), unspecified (55), other (3)
Respiratory 192 1⋅2Unspecified disease or failure (61), pneumonia (46), larnygeal paralysis (30), choked (14), bronchitis (11),
tracheal collapse (8), other (22)
Immune mediated§ 159 1⋅0Unspecified (88), AIHAe (42), ITPf (12), Evans’ syndrome (3), other (14)
Reproductive 144 0⋅9Pyometra (106), prostatic disease (18), other (20)
Poisoning 84 0⋅5Unspecified
Subtotal 15,392 96⋅9
Other∞489 3⋅1Other causes of death that did not fit into any of the above categories
Total 15,881 100
*Cancer: lymphosarcoma includes lymphoma and leukaemia
†Cardiac: MVD = mitral valve disease
‡Gastrointestindal: GDV = gastric dilatation/volvulus; HGE = haemorrhagic gastroenteritis
§Immune mediated: AIHA = autoimmune haemolytic anaemia; ITP = immune-mediated thrombocytopaenia
∞Other included the following reported causes of death in descending order: senility, internal bleeding, collapse, infection, ocular, dermatologic, por tosystemic shunt, died, septicaemia,
systemic, cardiovascular, pining, hyperthermia and drowning
and others 2003). The proportional mor-
tality of 26⋅9% for cancer in the present
study is higher than that found in the
Danish KC study (14⋅5%; Proschowsky
and others 2003) and the earlier UK
study (15⋅7%; Michell 1999), but it is
similar to that found in other studies
such as the USA necropsy study (23%;
Bronson 1982) and a German study of
life expectancy (27⋅3%; Eichelberg and
Seine 1996). The breeds with the highest
proportional mortalities for cancer in the
present study are similar to those breeds
that have been reported previously. The
overall proportional mortality of 17⋅8%
for “old age” in the present study is simi-
lar to that found in the Danish KC study
(20⋅8%; Proschowsky and others 2003)
and the earlier UK study (20⋅7%; Michell
1999). Deaths were coded as “old age” in
this study even when another condition
was reported since many of the additional
causes of death given were non-specific
and appeared to be a reason for eutha-
nasia, such as arthritis, incontinence or
colitis. It would have been useful to know
whether the dog had died or had been put
to sleep so that the causes of death and
euthanasia could be separated but, unfor-
tunately, this was not asked on the ques-
tionnaire. Although behaviour has been
reported to be a much more common
cause of death as a reason for euthanasia
in other studies (Patronek and Dodman
1999), behaviour problems had a very
low proportional mortality in this study
of 1⋅3% as well as in the Danish KC study
(6⋅4%; Proschowsky and others 2003).
There are many potential reasons why
behaviour problems were not a common-
ly reported cause of death. It may be that
in our study and the Danish study, own-
ers of pedigree dogs were more commit-
ted to continuing to care for their dogs
in spite of behaviour problems or it may
have been that the dogs were surrendered
for rehoming if they had behaviour prob-
lems. The median age at death due to a
behaviour problem was low at 4⋅58 years
and this is consistent with other studies.
Further work is continuing in the UK
using breed-specific surveys to gather
confidential health information on dogs.
There are many diseases, particularly
those complex diseases such as the vari-
ous forms of valvular heart disease, car-
diomyopathy, cruciate ligament injury
and immune-mediated diseases, that have
both genetic and environmental com-
ponents to their causation. These com-
plex diseases require a combination of
research efforts that include input from
veterinary surgeons in practice and in
research, epidemiologists and geneticists.
Research requirements for future disease
surveillance measures include the need
for clear case definitions of disease and
accurate and validated data collection
on the health status of clinically affected
and non-affected dogs that can be linked
to the pedigree for analysis and future
work on estimated breeding values. This
work must be done in cooperation with
the breed club health coordinators and
the KC.

V. J. Adams and others
522 Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association
Table 5. Overall and breed-specific proportional mor talities for the four most common causes of death in 72 breeds,
excluding the 828 (5.2%) of deaths due to unknown/unreported causes
Breed All
deaths
Cause of death
Cancer Old age Cardiac Urologic
N N % 95% CI N % 95% CI N % 95% CI N % 95% CI
Afghan hound 143 44 30⋅823⋅238⋅32920⋅313⋅726⋅91510⋅55⋅515⋅574⋅91⋅48⋅4
Airedale terrier 81 32 39⋅528⋅950⋅21113⋅66⋅121⋅067⋅41⋅713⋅178⋅62⋅514⋅8
Alaskan Malamute 14 5 35⋅710⋅660⋅8214⋅30⋅032⋅600⋅00⋅00⋅000⋅00⋅00⋅0
Basset Griffon Vendeen 76 25 32⋅922⋅343⋅51823⋅714⋅133⋅256⋅61⋅012⋅245⋅30⋅210⋅3
Basset hound 142 44 31⋅023⋅438⋅61913⋅47⋅819⋅0117⋅73⋅312⋅185⋅61⋅89⋅4
Beagle 241 79 32⋅826⋅938⋅73815⋅811⋅220⋅43012⋅48⋅316⋅683⋅31⋅15⋅6
Bearded collie 278 54 19⋅414⋅824⋅17225⋅920⋅731⋅0165⋅83⋅08⋅5227⋅94⋅711⋅1
Belgian shepherd 113 26 23⋅015⋅230⋅81513⋅37⋅019⋅565⋅31⋅29⋅432⋅70⋅05⋅6
Bernese mountain dog 394 180 45⋅740⋅850⋅6246⋅13⋅78⋅5205⋅12⋅97⋅2164⋅12⋅16⋅0
Border collie 106 25 23⋅615⋅531⋅71917⋅910⋅625⋅276⋅61⋅911⋅343⋅80⋅17⋅4
Border terrier 177 34 19⋅213⋅425⋅05430⋅523⋅737⋅352⋅80⋅45⋅3126⋅83⋅110⋅5
Boxer 130 50 38⋅530⋅146⋅82821⋅514⋅528⋅696⋅92⋅611⋅310⋅80⋅02⋅3
Briard 71 27 38⋅026⋅749⋅368⋅52⋅014⋅957⋅01⋅113⋅045⋅60⋅311⋅0
Brittany 28 5 17⋅93⋅732⋅0621⋅46⋅236⋅613⋅60⋅010⋅400⋅00⋅00⋅0
Bull terrier 209 48 23⋅017⋅328⋅72712⋅98⋅417⋅53014⋅49⋅619⋅13114⋅810⋅019⋅7
Bulldog/British bulldog 180 33 18⋅312⋅724⋅0168⋅94⋅713⋅03620⋅014⋅225⋅842⋅20⋅14⋅4
Bullmastiff 96 36 37⋅527⋅847⋅244⋅20⋅28⋅266⋅31⋅411⋅155⋅20⋅89⋅7
Cairn terrier 124 24 19⋅412⋅426⋅33528⋅220⋅336⋅1118⋅93⋅913⋅9108⋅13⋅312⋅9
Cocker spaniel/English
cocker 289 85 29⋅424⋅234⋅75017⋅312⋅921⋅7279⋅36⋅012⋅7155⋅22⋅67⋅7
Dachshund (all) 245 41 16⋅712⋅121⋅45321⋅616⋅526⋅83514⋅39⋅918⋅7124⋅92⋅27⋅6
Dalmatian 199 38 19⋅113⋅624⋅65427⋅121⋅033⋅3136⋅53⋅110⋅0105⋅02⋅08⋅1
Dandie Dinmont terrier 62 14 22⋅612⋅233⋅0711⋅33⋅419⋅269⋅72⋅317⋅023⋅20⋅07⋅6
Deerhound 287 54 18⋅814⋅323⋅33512⋅28⋅416⋅07024⋅419⋅429⋅462⋅10⋅43⋅7
Dobermann 100 26 26⋅017⋅434⋅62222⋅013⋅930⋅11515⋅08⋅022⋅033⋅00⋅06⋅3
English setter 384 126 32⋅828⋅137⋅57218⋅814⋅822⋅7277⋅04⋅59⋅6184⋅72⋅66⋅8
English springer spaniel 90 24 26⋅717⋅535⋅81617⋅89⋅925⋅755⋅60⋅810⋅344⋅40⋅28⋅7
Field spaniel 68 20 29⋅418⋅640⋅21217⋅68⋅626⋅734⋅40⋅09⋅322⋅90⋅07⋅0
Finnish spitz 42 9 21⋅49⋅033⋅849⋅50⋅618⋅437⋅10⋅014⋅937⋅10⋅014⋅9
Flat-coated retriever 610 331 54⋅350⋅358⋅2548⋅96⋅611⋅1386⋅24⋅38⋅181⋅30⋅42⋅2
French bulldog 71 27 38⋅026⋅749⋅368⋅52⋅014⋅911⋅40⋅04⋅111⋅40⋅04⋅1
German shor thaired
pointer 159 47 29⋅622⋅536⋅73018⋅912⋅824⋅963⋅80⋅86⋅710⋅60⋅01⋅9
German spitz/klein or
mittel 43 7 16⋅35⋅227⋅31023⋅310⋅635⋅9614⋅03⋅624⋅300⋅00⋅00⋅0
German wirehaired
pointer 41 10 24⋅411⋅237⋅5819⋅57⋅431⋅612⋅40⋅07⋅212⋅40⋅07⋅2
Giant schnauzer 39 16 41⋅025⋅656⋅5410⋅30⋅719⋅8615⋅44⋅126⋅712⋅60⋅07⋅5
Golden retriever 927 360 38⋅835⋅742⋅0 172 18⋅616⋅121⋅1505⋅43⋅96⋅8363⋅92⋅65⋅1
Gordon setter 157 46 29⋅322⋅236⋅42616⋅610⋅722⋅42314⋅69⋅120⋅274⋅51⋅27⋅7
Hungarian wirehaired
vizsla 15 7 46⋅721⋅471⋅9213⋅30⋅030⋅5213⋅30⋅030⋅500⋅00⋅00⋅0
Irish red & white setter 179 54 30⋅223⋅436⋅93419⋅013⋅224⋅7147⋅83⋅911⋅8126⋅73⋅010⋅4
Irish setter 451 123 27⋅323⋅231⋅4 101 22⋅418⋅526⋅24510⋅07⋅212⋅792⋅00⋅73⋅3
Irish water spaniel 95 53 55⋅845⋅865⋅81010⋅54⋅416⋅711⋅10⋅03⋅111⋅10⋅03⋅1
Italian Spinone 47 21 44⋅730⋅558⋅9714⋅94⋅725⋅112⋅10⋅06⋅348⋅50⋅516⋅5
Keeshond 104 33 31⋅722⋅840⋅72726⋅017⋅534⋅465⋅81⋅310⋅376⋅71⋅911⋅5
Labrador retriever 574 179 31⋅227⋅435⋅0 132 23⋅019⋅626⋅4427⋅35⋅29⋅4111⋅90⋅83⋅0
Lancashire heeler 30 8 26⋅710⋅842⋅5826⋅710⋅842⋅526⋅70⋅015⋅613⋅30⋅09⋅8
(continued overleaf)

Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association 523
Results of a health survey of UK purebred dogs
Large Munsterlander 69 17 24⋅614⋅534⋅8811⋅64⋅019⋅145⋅80⋅311⋅311⋅40⋅04⋅3
Leonberger 47 21 44⋅730⋅558⋅912⋅10⋅06⋅3510⋅61⋅819⋅500⋅00⋅00⋅0
Lhasa apso 84 15 17⋅99⋅726⋅02732⋅122⋅242⋅133⋅60⋅07⋅567⋅11⋅612⋅7
Manchester terrier 32 7 21⋅97⋅636⋅21031⋅315⋅247⋅313⋅10⋅09⋅239⋅40⋅019⋅5
Miniature schnauzer 214 46 21⋅516⋅027⋅03616⋅811⋅821⋅82511⋅77⋅416⋅073⋅30⋅95⋅7
Newfoundland 269 73 27⋅121⋅832⋅55219⋅314⋅624⋅04316⋅011⋅620⋅462⋅20⋅54⋅0
Norwegian elkhound 71 23 32⋅421⋅543⋅31622⋅512⋅832⋅3811⋅33⋅918⋅645⋅60⋅311⋅0
Norwich ter rier 56 12 21⋅410⋅732⋅21730⋅418⋅342⋅4712⋅53⋅821⋅235⋅40⋅011⋅3
Nova Scotia duck
tolling retriever 9 3 33⋅32⋅564⋅100⋅00⋅00⋅0111⋅10⋅031⋅600⋅00⋅00⋅0
Pointer 145 30 20⋅714⋅127⋅33524⋅117⋅231⋅164⋅10⋅97⋅474⋅81⋅38⋅3
Rhodesian ridgeback 183 56 30⋅623⋅937⋅33318⋅012⋅523⋅6116⋅02⋅69⋅552⋅70⋅45⋅1
Rottweiler 137 62 45⋅336⋅953⋅61410⋅25⋅115⋅375⋅11⋅48⋅842⋅90⋅15⋅7
Saluki/gazelle hound 132 47 35⋅627⋅443⋅81712⋅97⋅218⋅61914⋅48⋅420⋅421⋅50⋅03⋅6
Samoyed 223 59 26⋅520⋅732⋅24620⋅615⋅325⋅9114⋅92⋅17⋅862⋅70⋅64⋅8
Shetland sheepdog 364 81 22⋅318⋅026⋅55414⋅811⋅218⋅5287⋅75⋅010⋅46317⋅313⋅421⋅2
Shih-tzu 83 12 14⋅56⋅922⋅01720⋅511⋅829⋅21518⋅19⋅826⋅41315⋅77⋅823⋅5
Siberian husky 129 41 31⋅823⋅739⋅82116⋅39⋅922⋅686⋅22⋅010⋅443⋅10⋅16⋅1
Soft coated wheaten
terrier 111 29 26⋅118⋅034⋅32320⋅713⋅228⋅343⋅60⋅17⋅198⋅13⋅013⋅2
Staffordshire bull
terrier 117 52 44⋅435⋅453⋅42319⋅712⋅526⋅986⋅82⋅311⋅465⋅11⋅19⋅1
Standard poodle 118 35 29⋅721⋅437⋅92117⋅810⋅924⋅765⋅11⋅19⋅054⋅20⋅67⋅9
Sussex spaniel 42 8 19⋅07⋅230⋅9819⋅07⋅230⋅9614⋅33⋅724⋅900⋅00⋅00⋅0
Tibetan spaniel 125 29 23⋅215⋅830⋅63427⋅219⋅435⋅01411⋅25⋅716⋅743⋅20⋅16⋅3
Tibetan terrier 95 30 31⋅622⋅240⋅92122⋅113⋅830⋅499⋅53⋅615⋅488⋅42⋅814⋅0
Weimaraner 242 58 24⋅018⋅629⋅33313⋅69⋅318⋅02811⋅67⋅515⋅641⋅70⋅03⋅3
Welsh corgi Cardigan 53 15 28⋅316⋅240⋅41324⋅512⋅936⋅123⋅80⋅08⋅911⋅90⋅05⋅5
Welsh corgi Pembroke 116 33 28⋅420⋅236⋅72622⋅414⋅830⋅097⋅82⋅912⋅6108⋅63⋅513⋅7
Welsh springer spaniel 157 42 26⋅819⋅833⋅72918⋅512⋅424⋅574⋅51⋅27⋅7127⋅63⋅511⋅8
Welsh terrier 23 10 43⋅523⋅263⋅7521⋅74⋅938⋅600⋅00⋅00⋅014⋅30⋅012⋅7
Other breeds (n=93) 4524 806 17⋅8 811 17⋅9 808 17⋅9 268 5⋅9
Total 15,881 4282 27⋅0 2829 17⋅8 1770 11⋅1 783 4⋅9
Table 5. Continued
Breed All
deaths
Cause of death
Cancer Old age Cardiac Urologic
N N % 95% CI N % 95% CI N % 95% CI N % 95% CI
This study shows breed differences
in lifespan and causes of death, and the
results support previous evidence that
smaller breeds tend to have longer lifes-
pan compared with larger breeds. The
long-lived breeds died of diseases appro-
priate to their longevity with cancer, old
age and chronic renal failure represent-
ing the highest proportional mortalities
for these breeds. Although many of the
breeds in the study were not representative
for the breed club or the general pedigree
dog population in the UK, the results do
contribute to the limited information cur-
rently available on canine mortality.
Acknowledgements
The authors wish to thank the UK KC
Charitable Trust for funding this sur-
vey as well as the KC/British Small
Animal Veterinary Association Scientific
Committee and the participating breed
club secretaries and members for their
assistance in this study.
Conflict of interest
None of the authors of this article has a
financial or personal relationship with
other people or organisations that could
inappropriately influence or bias the con-
tent of the paper.
References
ALDERTON, D. (1993) Dogs. London: Dorling Kindersley
BONNETT, B. N., EGENVALL, A., OLSON, P. & HEDHAMMAR, A.
(1997) Mortality in insured Swedish dogs: rates
and causes of death in various breeds. Veterinary
Record 141, 40-44
BRONSON, R. T. (1982) Variation in age at death of dogs
of different sexes and breeds. American Jour nal of
Veterinary Research 43, 2057-2059
EICHELBERG, H. & SEINE, R. (1996) Life expectancy
and cause of death in dogs: the situation in
mixed breeds and various dog breeds. Berliner
und Münchener tierärztliche Wochenschrift 109,
292-303
EGENVALL, A., BONNETT, B. N., SHOUKRI, M., OLSON, P. &
HEDHAMMAR, A. (2000a) Gender, age, breed and
distribution of morbidity and mortality in insured
dogs in Sweden during 1995 and 1996. Veterinary
Record 146, 519-525
EGENVALL, A., BONNETT, B. N., SHOUKRI, M., OLSON, P.,
HEDHAMMAR, A. & DOHOO, I. (2000b) Age pattern
of mortality in eight breeds of insur ed dogs in

V. J. Adams and others
524 Journal of Small Animal Practice • Vol 51 • October 2010 • © 2010 British Small Animal Veterinary Association
Sweden. Preventive Veterinary Medicine 46,
1-14
GAROSI, L. S., MCCONNELL, J. F., PLATT, S. R., BARONNE, G.,
BARON, J. C., DE L AHUNTA, A. & SCHATZBERG, S. J. (2005)
Results of diagnostic investigations and long-term
outcome of 33 dogs with brain infarction (2000-
2004). Journal of Veterinar y Internal Medicine 19,
725-731
HAYASHIDANI, H., OMI, Y., OGAWA, M. & FUKUTOMI, K. (1988)
Epidemiological studies on the expectation of
life for dogs computed from animal cemetery
records. Japanese Journal of Veterinary Science
50, 1003-1008
MICHELL, A. R. (1999) Longevity of British breeds of
dogs and its relationship with sex, size, cardio-
vascular variables and disease. Veterinary
Record 145, 625-629
PATRONEK, G. J. & DODMAN, N. H. (1999) Attitudes,
procedures, and deliver y of behavior services by
veterinarians in small animal practice. Journal of
the American Veterinary Medical Association 215,
1606-1611
PATRONEK, G. J., WATERS, D. J. & GLICKMAN, L. T. (1997)
Comparative longevity of pet dogs and humans:
implications for gerontology research. The Journals
of Gerontology Series A: Biological Sciences and
Medical Sciences 52, B171-B178
PROSCHOWSKY, H. F., RUGBJERG, H. & ERSBOLL, A. K. (2003)
Mortality of purebr ed and mixed-breed dogs in
Denmark. Preventive Veterinary Medicine 58, 63-74
REID, S. W. J. & PETERSON, M. M. (2000) Methods
of estimating canine longevity. Veterinar y Record
147, 630-631