View report page

24
The persistency of this feature of the probability curve for the first year of life has led me to
make further analysis of the monthly mortality figures, and the following table has, therefore, been
prepared, which shows, for the quinquennium 1905-9, the deaths per thousand living in London
at the commencement of each month of life from certain groups of causes of death, and from " all
causes":—

An examination of this table shows that with regard to the common infectious diseases,
there is an almost uninterrupted increase in the rate of incidence from the first to the twelfth
month of life, while the reverse is practically true for premature birth, want of breast milk, etc.,
and "all other causes." The rates shown for tubercular diseases exhibit remarkable constancy
from the fourth month of life onwards, while the age-incidence curve of diarrhœal diseases rises to
a maximum in the fourth month of life, and then progressively declines. The age-incidence curve
for respiratory diseases is, however, much less regular, and further analysis of the figures is necessary
to ascertain the cause or causes which determine the shape of the curve, a shape which is suggestive
of increasing susceptibility to some forms of respiratory diseases and increasing resistance to others
as the age of the child advances. It will be seen that the marked increase in the incidence of these
diseases during the ninth month is sufficient alone to account for the depression in the probability
curve at this point in the London infant life tables. The figures for the whole of England and
Wales for the same period relating to respiratory diseases show a maximum incidence at the ninth
month, but the excess over the eighth month is not nearly so great proportionately as in London.
It seems, therefore, reasonable to conclude that an average over a greater number of years in London
will somewhat reduce the disparity between the respiratory-diseases deaths at the eighth and ninth
months of life. It is, however, clear from an examination of the death-rates from all causes in
England and Wales over a series of years that the influences operating in the first year of life are
such as to produce normally a flattening, hardly amounting to a depression, of the probability
curve about the eighth month, this being the natural result of the rapid decrease, down to this point,
of the deaths from pre-natal and natal influences on the one hand, and the increase from the sixth
month onwards in the deaths from infectious diseases and, to some extent, respiratory diseases on
the other hand.
Broadly speaking, the table shows that the causes of infantile mortality can be roughly
grouped into three classes:—(a) those causes increasing in incidence with progressive age, such
as the common infectious diseases; (b) those causes showing little or no variation in age incidence,
such as tuberculous diseases; and (c) those causes exhibiting diminished incidence as age advances,
such as premature birth, want of breast milk, etc.
It will thus be seen that the composite "all ages" curve is, in the main, a blending of two
curves, one approaching, the other receding from, the base-line, and an approximately straight line.
The contour of the "all ages" curve, therefore, will depend upon the relative steepness and distance
from the base-line of the auxiliary curves, which qualities must obviously vary with variations iu
the prevalence of the causes of mortality thus graphically represented. On these grounds alone, it
is easy to conceive of irregularities in the contour of the "all ages" curve similar to that to which
attention has been directed in connection with the probability curve. The expiration of a longer
period than that under review, and the consequent elimination of "epidemic influence," is necessary
to demonstrate whether the irregularity in the curve is of a purely arithmetical character, in which
case the irregularity will be smoothed out when a curve is constructed on a larger number of observations
1 See footnote (1), page 22.