Why de-sexed dogs live 2 yrs longer - critique of data analysis of 70,000 dogs!

The rigourus 2013 dog study used data for dogs that died in veterinary teaching hospitals between 1984 and 2004

NOTE - while it might appear that sterilisation increased the likelihood of cancers, it was emphasised that "sterilized dogs might have a higher occurrence of diseases that occur late in life (such as cancer) simply because sterilized dogs live longer." REF 1

Statistical analysis of results removed the effect of diseases that were most prevalent in old age dogs from confounding the results. Likewise diseases most prevalent in certain breeds were accounted for and bias removed from the results

Dog deaths were categorised into nine pathophysiologic processes (PP) : (congenital, degenerative, infectious, immune-mediated, metabolic, neoplastic, toxic, traumatic and vascular)

"The initial dataset contained 80,958 records of dog death. When juvenile dogs and those with unknown sterilization status were removed there were 70,574 FC dogs, representing 185 breeds. The average number of diagnoses recorded per dog was 2.9 (range 1–32). Overall, 30,770 (43.6%) dogs were intact and 39,804 (56.4%) dogs were sterilized at the time of death. The mean age of death for intact dogs was 7.9 years versus 9.4 years for sterilized dogs."

Effect of sterilization on longevity and diagnoses.

The statistical explanation of the four graphs below are given below the graph. My lay-mans interpretation is given immediately below:

Graph A shows that for males and females the BLUE LINE (sterilised dogs) is typically higher in years (longevity) than the red line- for total combined deaths. That is sterilised dogs generally outlive intact dogs.

Graph B The NINE PP (causes of death) that dogs can die of are shown for sterilised dogs. Any disease (the top ones on the graph) that are to the right of ZERO on the x axis mean that these pathological processes that caused death are more probable in sterilised dogs than intact dogs. The two major classes of sterilised dogs dying ahead of intact dogs is in:

Immune mediated
Neoplastic (a new, often uncontrolled growth of abnormal tissue; tumor)

The PP's below these two PP's (ie the other seven categories) are more prevalent in intact dogs (ie especially Infectious diseases and Traumatic pathology.

GRAPH C - the specific neoplastic diagnoses (cancers) most prevalent in sterilised dogs that cause death. The main ones are:

Transitional CC
Oesteosarcoma

You will note that by comparison to intact dogs death by mammary cancer (male or female) is much rarer in sterilised dogs.

GRAPH D - This is for sterilised dogs showing that the infectious diseases (that kill may more intact dogs) are relatively rare for sterilised dogs. In particular sterilised dogs have a much lower death rate from parvovirus, heartworm and canine distemper.

The statistics fine print:

(A) Kaplan-Meier plots of longevity for males (left) and females (right). Blue lines indicate sterilized dogs and red lines indicate intact dogs. (B) Common log-odds ratios with 95% confidence intervals (CI) for pathophysiological processes (PP). Height of each bar indicates the relative frequency of each PP among all deaths. (C) Effects of sterilization on specific neoplastic diagnoses, showing common log odds ratios and 95% CI. Height of each bar indicates fraction of individuals with this diagnosis at the time of death. Transitional CC – transitional cell carcinoma; Squamous CC – squamous cell carcinoma. All cancers significant atP<0.01 except prostate cancer, squamous cell carcinoma and melanoma (P>0.05). (D) As inFigure 1C, but for specific infectious disease diagnoses. All infectious diseases significant atP<0.01 except blastomycosis (P>0.4).