Discussion 17 



Benjamin: Two different kinds of error may be introduced by this 

 over-simplified analysis. I am more worried about my first assump- 

 tion that no one under the age of 55 can die of senescence. I am not 

 so much worried about this second assumption that no one over the 

 age of 76 can die otherwise than by senescence, because if an old 

 person dies of an accident, it is very difficult to know whether they 

 would have had the accident if they had not been old. In other 

 words, although a deaJth appears as accidental in the statistics it 

 may still be a death of senile origin. 



Danielli: To what extent can you correct your curves for acci- 

 dents ? There must be some proportion of accidents occurring to 

 which the person who was killed makes no contribution; for example 

 there are people who get killed in railway accidents, and as passengers 

 in cars. Then there are other accidents, such as those to car drivers 

 and motor cyclists, to which the individual concerned does make a 

 contribution. These two groups may show some variation with age 

 so that the non- contributory accident, so to speak, would be more 

 independent of age. 



Benjamin: You could calculate the deaths which are due to 

 accidents to which the individual may have made no contribution 

 so far as these are shown by the certified causes of death. You would 

 of course get rid of some part of that peak of "anticipated" deaths. 



Comfort: In the curve for deaths of pedestrians in road accidents 

 by age the mode is a very close fit with the curve of general deaths if 

 allowances are made for the different risks to infants. Pedestrian 

 deaths are an excellent measure of general vigour — the power to see 

 a vehicle coming, jump fast enough to avoid it, and recover if it hits 

 you (see Comfort, A. (1957). Ciba Found. Coll. Ageing, 3, 7. London: 

 Churchill). 



Benjamin: Isn't it true to say that pedestrians do make a contri- 

 bution to the accident ? 



Comfort: They make a large contribution and that is the point. 

 An inestimable amount of this pedestrian mortality is of an age- 

 distributed, or age-conditioned kind. The same point arises over my 

 horses, when I have to decide which I am going to call natural deaths 

 and which I am going to call non-natural deaths. It is a point where 

 the scoring convention becomes very difficult to determine. 



Jalavisto: Death from appendicitis might almost be described as 

 an accident. Mortality in appendicitis was formerly a nearly 

 horizontal line, especially in males, and it did not rise much with 

 age. Later on when conditions improved, it can be seen that it is 

 especially young people who escape death from appendicitis. The 

 result is that the curve rises and begins to resemble the form of the 



