There were 410 ferrets reported - 69 deceased, 341 living.  However 3 were
killed by bureaucracies or neighbor's pets and so were not used for
age-related analysis based on 'age of decease' - their cases would not be
representative of how long ferrets live.  Five were reported without age
data of any kind, and so weren't considered as a matter of age, either.
 
This quick preliminary report is based on simplistic analysis.  Here data is
divided into two groups, living and deceased, subject to exclusions already
mentioned.  There were 65 deceased records in the one group and 337 living
records in the other.
 
Regarding 'age of decease' :
 
- The overall average age at death is 5.39(65 samples).  The median age of
  decease is 6 years.
- The average age of US ferrets at death is 5.50(53 samples).
- The average age of non-US ferrets at death is 4.33(7 samples).
- The average age at death of ferrets whose country was unreported is 5.70
  (5 samples).
- The average age of albino ferrets at death is 6.42(6 samples).
- The average age of sable ferrets at death is 5.43(28 samples).
- The average age of bred-color (blaze, 'calico') ferrets at death is 5.32
  (28 samples).
- The average age at death of ferrets whose color was unreported is 3.64
  (3 samples).
- The average age of commercially-bred ferrets at death is 5.38
  (17 samples).
- The average age of privately-bred ferrets at death is 5.85(11 samples).
- The average age at death of ferrets whose breeding was unreported is 5.26
  (37 samples).
- The average age of female ferrets at death is 5.31(37 samples).
- The average age of male ferrets at death is 5.30(25 samples).
- The average age at death of ferrets whose sex was unreported is 7.25 (3
   samples).
 
I find the distribution of mortality by annual cohorts to be uncanny -
notice that the distribution is very regular (almost Gaussian but perhaps
rather 'Boltzman').  Most importantly it seems that half of the ferrets in
this group died between the age of 5 years and 8 years.
 
Deaths at age : 5     3     1     5    5    11    15    11     6    3      0
age (a)        a<1   1,2   2,3   3,4  4,5   5,6   6,7   7,8   8,9  9,10 10<a
 
Regarding the living records :
I am not currently capable of much useful analysis.  However, consider the
population reported, by annual cohort :
 
reports  :   81    79    53    38   38    21    10    13     3     1      0
age (a)      a<1   1,2   2,3   3,4  4,5   5,6   6,7   7,8   8,9   9,10  10<a
 
Again, it would appear that this group shows sign of steady pressure on
survivorship.  I have no idea how to interpret the median age of 1.0 years,
average age 2.50.
 
Overall health risks :
Adrenal disease and insulinoma were reported with 9.7% of these individuals,
lymphoma with 3%, and unspecified tumors with 1.7%.  Organ weaknesses
reported include heart (3%), liver (2%), and spleen (1.5%).
 
Conclusions : none.  I'm not smart enough to conclude anything.  However,
let me register some frustration - the reports were erratic regarding age.
Occasionally, a person would submit multiple reports, giving the same ferret
differing ages.  Also, ages were most often approximated (try identifying
your own cohorts in this group, and you'll see how ages 'clump' by year
first, and by half-year and quarter).  I have no rationale for identifying
an experimental error with these statements of age.
 
Because of the inexact nature of anecdotal reports, I think a more
controlled study is the only valid way to go.  Also, raw ages are not of
much interest - it is far better to be able to correllate a shortened life
with some factors, and again a very controlled study would keep track of
factors.
 
Such a controlled study, perhaps would identify owners and individuals in
advance of the inidividuals births.  The individuals would be followed
through life as in a so-called 'Framingham study'.  Rigorous records of
birth dates, weening, neutering, diet, disease, health care, injury and
trauma, reproduction, and environmental factors like play-time or types of
toys might be kept.
 
Now try doing that for 100,000 ferrets, and you can see how expensive it
becomes.  Perhaps that is why it's not done already.
 
NOTA BENNA : for those that really can analyze, the overall database is
available as a flat, 'comma-seperated-value' file by anonymous ftp from
quasimodo.libs.uga.edu - see pub/ferret.age/ferrets.csv.  Look for a full
report at that site later.
[Posted in FML issue 1920]