Wednesday 27 April 2011

Bigger is not always better

Did you know that no common mouse has made it beyond its fourth birthday while Asian elephants can live beyond 80? That’s because large mammals usually live longer than small ones.

Humans (actually all primates) are a bit of an anomaly - our maximum lifespan is longer than expected for our size and more than many bigger animals.  Frenchwoman Jeanne Calment, the oldest known human, managed 122 years and 164 days.
But oddest of all, perhaps, are dogs. And that’s because the rules are completely reversed: smaller ones live longer than big ones and the difference is extremely marked.
In the 2004 Kennel Club breed health survey, there wasn’t a single large or giant breed in the Top 14 of the longest-living breeds. At the other end of the scale, eight of the 11 breeds that lived less than 8 years were either giant or large breeds.
The giant breed with the highest median age-of death was the Newfoundland at 9.67, while the Great Dane could muster only 6.5yrs, the Dogue de Bordeaux only  3.83 yrs and – worst of all with an average lifespan of just  2.33yrs  - was the Neapolitan Mastiff (although this was based on very few survey returns – 90 were sent out and only 9  were returned, almost certainly skewing the data).
The longest living breeds included the Lakeland Terrier (15.46 yrs),  the Toy Poodle (14.63yrs)  and the Tibetan Spaniel (14.42 yrs) – all small breeds.  The oldest dog in the survey – at 22 years old – was a Border Terrier (a breed with a median age of death of 14).
It is not completely linear - quite a few large-ish (if not giant) breeds do muster a respectable lifespan. Border collies, Dalmatians and most spaniels often make it to 12.5yrs. And there are some exceptions in the small-dog category:  the Japanese Chin averaged just 9.25yrs, the Pomeranian 9.67yrs and the Cesky Terrier just 8.42 yrs - half that of many other terrier breeds (and probably reflecting the very high level of inbreeding in the Cesky).
The KC survey had its limitations: it wasn’t random (so may not be representative of the breed as a whole) and some breeds returned very few survey forms. The oldest Leonberger in the KC survey for instance, was 12.67 (from a sample of around 200 dogs), whereas the excellent International Leonberger Database now has data on more than 3000 Leos and the eldest recorded died at 17.
But, overall, the evidence is pretty strong and every other survey has found much the same:  large dogs die younger than small ones. Given the reports of some very big dogs living well into their teens, it’s not that they’re incapable of living longer. It’s that something causes them to die prematurely.
But what? There are two main theories. The first is that the bigger and heavier the dogs are, the greater the strain there is on the skeleton and vital organs such as the heart that have to work that much harder. In other words, bigger breeds simply wear out more quickly.
And if that’s not true of elephants, it’s because elephants have been honed for millennia by natural selection to be the size they are.  In contrast, many dog breeds are only 200 years old or less. They may have evolved so rapidly thanks to human selection that it’s possible some body systems have had a hard time keeping up.  There is some evidence of this in the mismatch between skull and brain size that is thought responsible for the neurological problems in Cavalier King Charles Spaniels.
Elephants also grow much more slowly than dogs. And this ties in with the second theory – which is that giant dog breeds die younger because they grow too fast. Clearly, small breeds mature a bit quicker than large ones but the difference is not that great given the huge difference between a Chihauhua and a Great Dane (the equivalent, incidentally in human terms of some adult humans being just 2ft tall and others 31ft!)
Giant breeds suffer more than their fair share of problems associated with both this rapid growth and from being so darn heavy. These include joint problems such a hip dysplasia and bone cancer. The bone cancer often seen in big dogs is the same type that hits human adolescents and it is strongly associated with the teenage growth-spurt. Many breeders of large and giant dogs are aware of the problem – and try to limit it by feeding a lower-protein food and limiting exercise, but there may be more they should be doing.
In a 2007 paper, Dutch scientist Frietson Galis asked the question: “Do large dogs die young?” (abstract here)The answer was yes overall. I asked her if she thought we should be breeding smaller dogs. “We think that health could improve by selecting for slower growth, she says. “The problem is not so much the size, it’s that they have selected for the fastest growing individuals.”
Surprisingly, Galis found from one large database (although not another) that smaller individuals within a breed lived very slightly less long than larger ones – the reverse of what one might expect. given what we know about dogs. But I still think we should be selecting for smaller big dogs.
Frietson Galis worked out that Great Danes increase in weight 100-fold in their first year, compared to 60-fold in captive wolves , 20-fold in Poodles and just 3-fold in humans.  Clearly, if Great Danes weren’t so big, they wouldn’t have so much fast-growing to do.
Some breeds still doing the job for which they were bred – such as livestock guardian breeds – do need to be a certain size. But in others there really is no good reason, beyond human whim, to breed bigger and bigger dogs.
The problem is that bigger is often seen as better in the dog world, particularly in the show ring.
I grew up with Great Danes – always big dogs, but the Danes in the ring today are much bigger than the dogs I knew and loved in the Sixties and Seventies. The same goes for many other breeds.  
The Berne Natural History Museum in Switzerland has extensive data on the St Bernard dating back to the 1800s. The UK breed standard for the St. Bernard now specifies a minimum shoulder height of 75cm (30in) for dogs and 70cm (27.5in) for bitches and they weigh 65-85kg (143-187lb). But a typical 19th century dog was approximately 60 cm (20in) high and weighed less than 50 kg (110lb) . It is a huge increase.
“Breeds like the Bullmastiff, the Dogue de Bordeaux, the Neapolitan Mastiff and the Mastiff were never intended to be admired and valued because of their size but for what they could do,” says canine historian David Hancock, who has a particular passion for the Mastiff. “No breed of dog benefits from being too big; that is a human desire for dogs to suit their concept rather than the dog's best interests.”
David is dismayed by the fashion for “good bone” as it’s called.  “Strength, power and endurance do not reside in heavy bone, as any dog-sledder will tell you. To breed dogs with bone heavier than nature intended is asking for trouble, as the statistics on hip and elbow dysplasia, cervical vertebral malformation and osteochondrosis sadly demonstrate.”
David recalls the following boast made by a Mastiff breeder advertising his stock in the dog press a fear years ago: 'I am pleased to say that we have now bred the largest and heaviest dog in Britain'. The dog in question was 89cm (35in) tall and weighed over 127kg (275lb).
Unfortunately, some breed standards have encouraged excessive size too, with minimums but not maximums set for some of the giant breeds.  Until the revision of breed standards that followed Pedigree Dogs Exposed, the breed standard for the St Bernard demanded: “Taller the better”.  It has now been changed to: “Size is desirable but only if combined with quality, correct balance and absolute soundness.” Again, it stipulates a minimum height, but not a maximum.
The breed standard for the Deerhound, too, sets a minimum desirable height but no maximum, while the standard for the Great Dane sets both a minimum height and a minimum weight that the dog must have achieved by the age of 18 months (a stipulation that positively encourages rapid early growth).
As recently as 100 years ago, some accounts of Newfoundlands state their average weight at around 45kg (99lb),” says canine historian Scottie Westfall, who writes the Retrieverman blog. “It was a large dog, but not today's giant Newfoundland that can exceed 70kg (154lb).” In fact, the largest Newfie on record weighed a whopping 120kg (264lb).
The breed standard for the Irish Wolfhound even includes this:  “Great size, including height of shoulder and proportionate length of body is to be aimed at, and it is desired to firmly establish a breed that shall average from 81-86 cms (32-34 ins) in dogs.”
It takes a big, tough dog to kill a wolf in a battle,” explains Scottie Westfall. “In proportion to their body size, dogs have smaller heads and teeth than wolves do, so it was necessary to breed really big dogs if someone wanted them to fight wolves. This is probably the main selection pressure that drove the Irish wolfhound to its great size. If we are to believe saga and ancient texts, Irish wolfhounds were originally larger dogs than they are today, and their fighting prowess was renowned.”
Taking the old texts as gospel, then, the breeders have decided to try to breed an even bigger Wolfhound. Given what we now know about the cost, perhaps it is time to re-think this.
Over the years, some breed standards have even upped the minimums, too, as the dogs have got bigger and bigger (the breed standard for the Clumber Spaniel has seen no less that three weight revisions  - all upwards).
Scientists now know that a surprisingly few number of genes govern the huge differences in different breeds of dog. In April 2007, researchers announced that just one gene – called IGF1 – determines a breed’s average size (although other genes and environmental factors account for why one German Shepherd will be a bit bigger than another).
More recently, another gene - HMGA2 – has been found to control a dog’s height with different variants of this gene dictating whether a dog is short, medium or tall.
There are many other things beside size that impact on longevity – including diet, environment , gender and also whether or not dogs have been neutered.  A fascinating study of Rottweillers published in 2009 found that female rotties were more likely than males to be exceptionally long-lived – but not if they had been spayed in their first four years of life.
Clearly, there is much more to unravel.  But given what we know about size and longevity, I believe we should be cautious about growing bigger and bigger dogs. It would be good to see Kennel Club and breed clubs to get together to bring in more limits. In dogs, size really does matter – and bigger is not always better.
This article appears in the April  2011 issue of Dogs Today magazine.  Dogs Today is now available internationally for iPad and iPhone for the bargain price of 59p for the app, which includes one edition free.


  1. Very interesting article. Thanks

  2. Maybe we should imitate the American Dobermann Club, and breed for longevity in all KC-recognised breeds! See under following link:

  3. You can't really breed for longevity until you understand what the dogs are dying of. DCA have their longevity award because dobermans are well known for kicking the bucket due to cardiac incident which is rampant within the breed among a litany of other issues. It says much when they have to give an award for a dog who manages to beat the odds and live past ten when dogs of similar size regularly live into their mid teens in pretty decent health. That's not longevity, that's just luck.

    There isn't anything I admire the DCA for. They have ruined a great breed for pursuit of a better showring dog even though the dog's purpose is still high in demand.

  4. Oh dear Jemima! You've fallen into the same trap as your friend Mr Burns.

    Statistics can be used to prove most things - but not all of them are true.

    "the Cesky Terrier just 8.42 yrs - half that of many other terrier breeds (AND PROBABLY REFLECTING THE HIGH LEVEL OF INBREEDING IN THE CESKY)."
    (my emphasis)

    If your post is supposed to be scientifically accurate, the Cesky Terrier is not a good example to use in this instance - despite giving the opportunity for a jibe about inbreeding.

    The actual statistics were taken from a survey that covered just 13 owners, reporting on 22 live dogs and 9 that had died. Hardly a statistically representative sample!

    As the survey was carried out in 2004, and very few CT were bred in the UK before the late 1990s, it was statistically impossible to give an accurate figure for age of death at that time - most of the dogs of this breed registered in the UK were still alive!

    The ages of the dogs whose death has been reported to me as Breed Health Co-ordinator over the last 12 month period have ranged between 16 years 2 months and 7 years 10 months, with all but two reaching double figures, and an average age at death of 13 years 11 months. This is a rather more accurate figure.

  5. The surveys you cite do certainly seem prone to flaws... This retrospective data mining study just came out:

    75,000 dogs, 82 breeds, deaths broken down by process and organ system, with correlations to breed standard weight and age at time of death.

    There's still room for error here as the authors point out, but it seems more robust than a survey with a 10% return rate. I'd love to get my hands on that data set.

  6. There is a big qualification with the Fleming study, too, though - it's only vet school data (ie dogs that have been referred on by general pratices), so although useful, may not be representative of the breed as a whole.

    Great news is that there are plans afoot to give us a much better system of data surveillance - at least in the UK. Will be blogging soon.


  7. The giant breeds have certainly got far bigger in my lifetime. What seems to have happened is that the larger they get the less energy they have. I cann't imagine today's St Bernard being able to rescue anyone out in the snow, it's much too lumbering and ungainly.
    A young leonberger was trying to play with my spaniel the other day, and it was pathetic watching it crashing around, unable to chase her or do anything resembling proper play, other than a play bow.
    The Great Danes are absurdly big now too.
    Re the lifespan, I've noticed that it is the dogs that are lighter built, with smaller heads, that live the longest.
    They can still be medium sized, ie the whippet or the border collie, which both live a reasonable length of time, as it seems to be not length of leg but heaviness of bone and size of head that dooms some breeds to a shortened life.
    Julia Lewis

  8. The dog guarding the Hallaton treasure seems likely to have been unusually large for 1st century Britain and he was roughly the size of a modern GSD.

    It would be interesting to know whether there is any difference between breeds whose size has changed very fast over the past 200 years - I'm thinking of the Pomeranian which has shrunk from near Samoyed size vs terrier types.

  9. The larger 'Pomeranians' still exist as the German Spitz, American Eskimo Dog, and Italian Volpino. The breed name described a wider variety of spitz-type dogs originally than it does today -- it's less a matter of miniaturization than it is of the definition (and standard) of the breed changing over time.

    A good site that explains the diverse origins of the modern toy Pom is here:

  10. The Tibetan Mastiff is a generally long lived Large breed with a decided slow growth pattern, which would seem to uphold the data. Unfortunetly there are breeders trying to make them bigger, so I don't know how long that will last.

    1. Well, i agree, there are one or two breeds that can attain a giant size in world healthy, but even so tibetan mastiff and others shouldn't never surpass the line of 140(63 kg.) pounds if we want really working and long life dogs

  11. subaruthie, actually the KC only published the results of those breeds with at least 15% return rate. The average return rate is about 25%. I presume Jemima got the Neapolitan Mastiff stats directly from the KC.

    Unless you are asking for raw data, the summary for individual breeds can be found at

  12. Interesting. I think the nub of it is that human selection for extra large or extra small versions of an animal will mean a shorter lifespan. This is seen even in humans where dwarfism and giantism are linked to shorter lifespans. Natural selection occurs more slowly and fits the animal for it's environment. Certainly size seems to be shifting in many dogs breeds; the breed standard for the english springer calls for a dog around 20inches tall. My working dog is this height so why do the show dogs tower over him? There is also a trend in pet dogs to breed for 'teacup' examples; bizzarley tiny yorkies, jacks, shi zhus etc.

  13. On my dog family beach walk this morning I had another thought...does it matter how long a dog lives? Or is it more important how well a dog lives? If a Great Dane has 8 good years then dies after a short illness is that good...or a problem?
    I've seen many animals over the years of 15, 17 or 20years that the owners are deeply proud of but that I feel outlived their meaningful lives many years previously.

  14. Just FYI, Patricia McConnell opened a discussion of inbreeding on her blog

    Good to see so many great & innovative minds in the dog world considering this topic.

  15. Yes Vicki it matters. Don't they say a great dane has two years as a puppy, two years as an adult dog, and two years as a senior? For me, that is not even close to long enough. The payoff for dealing with a great dane puppy for two years should be more than two years of enjoying the mature adult dog before you have to deal with the diseases and problems of old age.

    I find it rather horrifying that you seem to be suggesting we just get used to things the way they are.

    I have an anatolian shepherd, not as large as a great dane but nevertheless a giant breed, who will celebrate his 13th birthday on June 13. It IS possible to have long-lived giant breed dogs--we just need to change the way we breed dogs and the way we think about breeding dogs.

  16. Sorry to 'horrify' you Kate...I was just putting a thought out there for debate. Extending he life of a great dane (for example) is fine but we must ensure it's worthwhile. There would be no point Danes living to 13 if they were still getting 'old' at 6. Does a 15 year old doubly incontinent, toothless yorkie which has to be carried everywhere really have a long and happy life?

  17. On the Yorkie mentioned above, maybe not, but my dogs (Border Collies, Papillons) were all very active and playful well into their teens. My current 13 year old Papillon runs and plays like the younger dogs and you could never tell by how he acts hes 13. I have never seen a giant breed over 5 act frisky at all. I knew a kick-ass Newfie who did search and rescue and she had to be retired at 6 (after 3+ years old training) as she was slowing down, she was gone by age 8. Kinda sad.

  18. Oh yes, a long, active happy life is best....but second best is a shorter, active happy life! Any attempt to increase longevity MUST also ensure extended QUALITY.

    I couldn't have a dane, or a bernese - I'm too selfish to risk a breed with a shorter than average lifespan, but I know many who do and would rather have the breed they love for a short time than switch to another breed. That doesn't mean they wouldn't rather have the same breed around for longer. Maybe good breeding can help? Maybe we can't help these breeds, but hopefully we can stop other breeds falling into the same trap.

  19. It wouldn't be that difficult to find a healthy a female from one of these breeds and breed her to the oldest, still healthy male you can find...It would be a start.

  20. @Kate: I can see your line of thought, but if old male humans are anything to go by, sperm quality of an old dog would be seriously compromised. Possibly better to go with the son of a long-lived dog.

  21. I think it's a red herring to talk about "quality of life" with increased longevity when we are discussing *breeding* practices.

    I know of no way that one could practice genetic selection for longevity that would not also improve life-long health.

    Now, when discussing *veterinary* protocols, quality of life issues are relevant. It's possible to use medical intervention and technology to prolong a life not worth living, as humans in rich developed countries are discovering to their dismay.

    But I challenge anyone to provide a reality-based scenario in which genetic selection for longevity would lead to animals that are decrepit geriatrics for five times longer than normal, which appears to be what some have invoked.

    Longevity is priority in my breeding program. Witness Beth F.'s account of the search and rescue Newfie retired at six. A long-lived dog would have had a ten-year (or longer) career, not a three-year career. That's seven years robbed from human beings who need a dog to save their lives. (It's one reason that we don't include giant breeds on our checklist of eligible-for-SAR dogs in our unit.)

    Funny thing by selecting for longevity (using mature bitches, old studs, watching longevity of ancestors and factoring that in) I am also, shockingly, selecting for robust good health, and *failing to select* for precocity, which may be linked to short life.

    Damn thing about a breeding program that puts a premium on longevity -- it sure takes a long time to show results. Good thing I had progenitors in this project and wasn't starting with Doberman-level genetic material.

    Last I looked, the DPCA's longevity hall o' fame included any Dobe that lived past age TEN. That is appalling. The Doberman is a breed in need of genetic rescue by judicious and fairly extensive cross-breeding. The resistance to that truth by Doberman "fanciers" is just as appalling as their dogs' ill-health and attenuated lifespans.

  22. One definitely *can* breed for longevity without knowing the causes of mortality. One might be able to do so more accurately knowing the causes of mortality more intimately, but until a century ago, virtually all breeding was based solely on function, not on the perceived existence of causes of those traits.

    So, I disagree that it makes sense to breed smaller big dogs if we want longer lived big dogs. Actually, what makes more sense is to breed *longer-lived* big dogs. Whatever it is that makes them live longer, be it small size or some other X factor, will be favored.

    Breeding for size on the assumption (a well founded one, I admit) that smaller size is healthier is an *indirect* way to get health. It's like breeding a supposed 45 degree shoulder angle in Dobermans as a way to get to a structure that is powerful and holds up for a long time. What about just breeding the dogs that ARE powerful and DO hold up for a long time? Let nature take care of the angle -- that's how evolution works. Same issue here.

    I don't disagree with the thrust of this blog at all -- the plight of the giant breeds is appalling to me. I'm commenting only about the sense of how we get there from here. We should absolutely do modern health testing, most especially for invisible or late-appearing issues, to avoid perpetuating particularly destructive genotypes or phenotypes. But past that, if we breed function rather than the form we hope/assume leads to function, we are taking a much more direct line to health.

  23. Great points, Greta - and, yep, you're right of course. If you want long-lived, healthy dogs... breed from long-lived, healthy dogs. Although I do think that may well mean breeding smaller ones.

    Thank you for commenting.


  24. You're very likely right, Jemima... And likewise, if we need Irish Wolfhounds to hunt wolves, then we need to use them to hunt wolves (ouch, I know, that's sensitive!) -- and let their success tell us how big they really need to be to do it well. Eh?