A few days ago, there was another of those news reports that say “If you do or don’t do this/eat or not eat this/indulge or avoid this … (insert your own favourite topic), you will live longer. Else you will suffer a premature death.” In this case, the directorate entreated us to eat ten pieces of fruit and vegetables a day, instead of five. If we all did this, the report said, 7.8 million premature deaths could be prevented every year. Quite apart from the ambiguity of the advice—do we eat ten pieces of fruit and ten pieces of vegetables a day (equal twenty pieces of fruit and veg a day) or a mix of ten pieces of fruit and vegetables a day (equal ten pieces total per day), or what?—I was somewhat intrigued by the notion of premature death. What is meant by premature death? Google defines premature death to be “death before a person reaches an expected age”; the insurance industry as “death that occurs before the stage where it is accepted by society as part of the natural, expected order of life” (whatever that means). But what is meant by expected age, expected order of life? Clearly, a young man in his twenties who is killed in a road accident suffers a premature death. Or does he? If the young man is in a city where the roads are congested, the drivers manic, and the number of fatalities higher than average—Phnom Penh at rush hour for example—then the young man did not suffer a premature death.

Bell-shaped distribution of life expectancy in the UK if you are 65 years old.
Snagged from http://theukpensionsblog.co.uk/tag/life-expectancy/

Premature death is normally defined by life expectancy statistics compiled by the life insurance industry. For example, the life expectancy of a male in the UK currently aged 75 (my age) is 86.26; for a female aged 73 (my wife’s age), it’s 93.87. (I know; life favours the females. They can have babies!) Consequently, if I die before I reach the grand old age of 86.26, will I have suffered a premature death? Not on your life for life expectancy is not a single value. It’s a point on a life expectancy curve. If the distribution is bell shaped (Gaussian), the peak is the average (arithmetic mean) value but we are all spread out somewhere on this distribution. Some will die earlier than the average value; others later. Some will die considerably earlier: through disease, through recklessness, through unhealthy life-style practices (heavy smoking, drinking, drugs, etc.), through accidents, and so on. Some will continue well beyond their expected average year of death. No; when you think about it, a single-point life expectancy is a meaningless concept, a statistic that reduces a huge spread of possibilities to a single point value. What then is meant by 7.8 million premature deaths that could have been prevented if we all consumed twice the recommended amount of fruit and vegetables a day? If these deaths were prevented, the life expectancy curve would change to reflect this and then the average life expectancy would be revised upwards thus causing yet more to die before their time thus increasing the number of premature deaths.

It makes no sense to me.

(^_^)