Well, except that it excludes some important considerations (and several of these effects are for large breeds or are breed specific). I have always thought NIAIA irresponsible in presenting one side of an issue, excluding what I think in the real, pet-owner world is far more likely to *actually* kill their dog, as opposed to incremental increases of mostly infrequent conditions.
First the 'risks' of many of these prospective conditions are extremely low to begin with so an increased risk is actually still miniscule over the life of the dog. Set that against the risk of not spaying a female cavalier: pyometra in cavaliers. This infection is frequently fatal and painful, coming on fast with no noticeable advance symptoms; and at best, very costly and risky to address with an emergency spay. According to a large Swedish study across dozens of breeds, the incidence of pyometra in cavalier females was amongst the very highest in ALL breeds examined with more than a third having pyo if left unspayed. Mammary cancer is also the most common cancer in female dogs, accounting for more than half of all tumours, and half are malignant. In rescue, we frequently see it in older unspayed female dogs. Waiting til after the second heat to spay still leaves a female dog with a one in four chance of this cancer.
But the article fails to look at the welfare/behavioural-related arguments -- which would include death rates for dogs that run far higher -- easily! than any of the possible medical risks.
How many of the unneutered dogs produce litters that end up at pounds, or are put down by the owner (numerous dogs dead in one go)? How many unspayed females die giving birth or after from complications (birth is after all a significant mortality risk). How many unneutered dogs (M&F) stray because of hormones calling? How many of those meet a sad end in accidents or are put down as strays in pounds? How many dogs end up surrendered to pounds or rescues or are just taken to the vets and requested to be pts for behavioural issues directly related to not having been neutered and where the owner has done no training to manage the issues (especially with males)?
On the behavioural side for males, from a UC Davis Vet School study, neutering brings: a 94% reduction in roaming -- 66% reduction in mounting -- 63% reduction in inter-male aggression -- 59% reduction in urine marking. Those are many of the main reasons people get rid of their dogs! I worked in general rescue in kill pounds where unneutered males always made up about 70% of all pound dogs; unspayed females (often in heat), another 25% or so. Neutered males or females make up only a tiny proportion. Most of those pound dogs are put down. In Ireland, that totals over 16,000 annually in pounds, mostly healthy dogs, being put down. It is millions in the US.
I would think NIAIA posting stuff like this would frustrate many reputable breeders (as I am sure it does). They don't seem to offer a counterbalance defending neutering -- about why reputable breeders sell puppies on spay/neuter contracts, why controlling indiscriminate breeding in pedigree dogs is essential, and what the welfare-related behavioural arguments for neutering are. It is far too easy to overlook the author's own (unfortunately, dismissively made) point:
This article will not discuss the impact of spay/neuter on population control, or the impact of spay/neuter on behavior
which are of course, both directly related to pet fatalities, and surely more significantly associated with risk of death than any of the health aspects reviewed. It is far too easy to scare the typical pet owner into thinking the statistically small increased risks cited here are major risks for their pets, so the message comes across that they should not spay at all.
PS I also wonder what someone else than the author might make of her same review of literature. A while ago, I went and read through some of the studies cited in this article and found that in several cases, the studies were few, the study samples small and the 'increased risks' were actually tiny. These points are not noted (ie a fourfold increase to a teensy percentage is a raised but not a significant risk; the article does not convey such critical points).
I've done doctoral level research myself. I would wonder at submitting an article like this without a more nuanced introduction about the counterbalance of the population and behaviour elements and overall death/accident risk, given the sensitivity of the topic and the misuse that could be made of it.