Jan 8, 2011

The Life Expectancy Variance Monster

After 'age', what would be the most important explanatory factor with regard to mortality rates or constructing life tables?

As actuaries we've demonstrated our innovation capabilities by developing life tables not only based on 'age' and 'gender', but also (two dimensional) on 'time', 'generation' and 'year of birth'. This helped us to extrapolate future mortality rates in order to predict future longevity with more accuracy.

However, despite our noble initiatives, these developments turn out to be insufficient to put the Longevity Variance Monster back in his cage.

Modern 'life expectancy at birth' predictions for periods of 40 to 50 year ahead, lead to 95% confidence intervals of 12 years or more. Unusable outcomes .....

Let's not even discuss more necessary accurate confidence intervals of 99% or more ....

In our attempt (duty?) to moderate and diminish future life expectancy variance, we'll have to develop new instruments.

The more we know which risk factors 'are responsible for the increase in 'life expectancy', the better we can estimate and diminish future variance.

One of those new approaches is to calculate life expectancies on basis of postcodes.

This new insight can be helpful, but there's a much more important risk factor that has to be included in our life expectancy predictions to definitely kill the Longevity Variance Monster:

Self-perception of aging

In a 2002 research "Longevity From Positive Self-Perceptions" by Levy ( et al.) it became undeniable clear that:
  • negative self-perceptions diminish life expectancy;
  • positive self-perceptions prolong life expectancy.
Older people with more positive self-perceptions of aging, measured up to 23 years earlier, lived on average (median survival) 7.6 years longer than those with less positive self-perceptions of aging. This advantage remained after age, gender, socioeconomic status, loneliness, and functional health were included as covariates.

Top 6 Life Expectancy Risk Factors
Here's Levy's top 6 list of risk factors on life expectancy (ordered from greatest to least impact on life expectancy):

  1. Age
  2. Self-Perceptions of aging
  3. Gender
  4. Loneliness
  5. Functional health
  6. Socio-economic status

As we can not change 'age' nor 'gender', let's put some more research on the other risk factors.

Once we achieve to 'explain' the cause of increase of life expectancy on basis of 'new' (soft) risk factors, we - as a society - will also be able to manage life expectancy better (information, education, training, coaching, etc.).

In this way actuaries can help society so that people live longer and stay happy in good health. All on basis of of a sound financial pension and health system, as predicted life expectancy will show a smaller variance.



Help to kill the Life Expectancy Variance Monster.....

Happy 2011, with better expectations and smaller variance!

Sources/related Links:

- Why population forecasts should be probabilistic
- On line Postcode Life Expectancy Tool
- Longevity From Positive Self-Perceptions
- Predicting successful aging (2010)

No comments:

Post a Comment