Life Expectancy

By: Michael P Barnes, MD

Life expectancy is a complex subject. What do we mean by life expectancy? A life expectancy of a cohort of a given age is the arithmetic mean of the actual survival time of the entire cohort. A more useful term is ‘median survival time’ which is the middle survival time for a cohort, if all the survival times are arranged from smallest to largest. In other words this is the time when 50% of the cohort has died and this is probably the most useful statistic.

Neither of these terms is actually the same as the actual survival time of an individual which can naturally vary significantly from the median value of the group as a whole. There are a multitude of factors that can affect life expectancy in the able-bodied population and even more that can have an effect in the disabled population. In providing an estimate of life expectancy the fact that the general population is living longer decade by decade needs to be taken into account. Life expectancy has increased by around 2.5 years every decade for at least the last 160 years.

There is, so far, no sign that this improvement is slowing down. The reasons for this increase are undoubtedly complex. In the early part of the 20th century the reasons for improvement in life expectancy were probably related to a reduction in malnutrition and improved treatment of infectious disease. In more recent years it is likely that continuing improvements in medical care are still having a positive effect combined with general improvements in both social care and income levels. Such positive elements will apply to the disabled population as well as the able-bodied population.

There is also an increasing awareness of the negative effects of various lifestyle factors such as smoking, obesity, excessive alcohol, lack of physical exercise and poor diet. There is now a reasonable body of literature that allows for an estimate of life expectancy for many of these lifestyle factors. It is important to emphasise that the disabled population is just as prone to such negative lifestyle factors as the able-bodied population and, indeed, there is evidence that some of these factors, including obesity, alcoholism and lack of physical exercise, are somewhat more common in people with disabilities. This certainly emphasises the need for good general health advice for disabled people.

However, people with disabilities are prone to a number of additional factors that are known to reduce life expectancy. The literature is clear that in non-progressive neurological disability (mainly traumatic brain injury, spinal cord injury and cerebral palsy) it is the disability itself that determines life expectancy rather than the underlying cause. There are five key factors that reduce life expectancy. The foremost is immobility. The risks associated with immobility are increased liability to pressure sores as well as an increased risk of osteoporosis, which in turn leads to a risk of limb fractures. Immobile individuals are also more prone to respiratory and renal problems.

A number of papers have demonstrated a graduated effect according to the degree of immobility. Individuals who can walk, or even stand in a non-functional sense, have a much improved life expectancy compared to those who are immobile. A paper by Hutton and colleagues1, for example, showed that 30-year survival in children with cerebral palsy was 97.5% for those with some mobility compared to just 42% for those who were wheelchair dependent. A similar finding in the context of spinal cord injury2 was better survival for those with incomplete paraparesis and are able to walk, even to a limited extent, compared to those with quadriparesis who are unable to do so.

The second key factor is incontinence. The risks of incontinence obviously include urinary infections with consequent risks of septicaemia as well as bladder stones and longer term kidney problems. Frankel and colleagues3 have recently shown that urinary deaths now only rank fourth after deaths from respiratory disease, heart disease and injury and there is no doubt that long term urological follow up and the advent of some techniques, such as intermittent self-catheterisation, have significantly reduced the risk of urinary complications. Such risks, particularly in those with spinal cord injuries, can be kept to a minimum by careful long term monitoring.

The third key factor is related to swallowing difficulties. A number of papers indicate that individuals who require tube feeding are at a risk of earlier death compared to those who are able to feed themselves. This may, at least partly, reflect the degree of underlying brain damage. The risks associated with swallowing difficulties have been significantly reduced by the advent of PEG feeding which is clearly superior to longer term nasogastric feeding, which should no longer be practised.

The fourth factor is epilepsy. There is a small risk of dying during a fit and a slightly higher risk of dying in between a fit in the so-called Sudden Unexpected Death in Epilepsy (SUDEP) syndrome. However, the risks associated with epilepsy are largely related to seizure type (generalised seizures having the highest risk) and seizure frequency. Better anti-convulsant drug regimes have improved the management of epilepsy which not only has made for better quality of life but has also improved life expectancy.

The final factor is the degree of cognitive and intellectual damage. This probably is a simple marker of the degree of underlying brain damage. A number of papers, for example, have shown that individuals with cerebral palsy with mild learning disability have virtually normal life expectancy but those with severe or profound intellectual damage are still at risk of earlier death.

These five factors are the most important with regard to life expectancy but other factors clearly may play a role in specific individuals. The presence, for example, of an intracranial shunt or severe scoliosis may well have an impact. Individuals with severe behavioural problems may also place themselves at risk. Chronic depression is regrettably common in the context of traumatic brain injury and spinal cord injury and will also carry a risk of suicide.

There are also more subjective factors, such as the quality of nursing care and environmental setting. Is, for example, an institutional setting better or worse for people with severe disabilities than living at home? Some authors have found an increased mortality in the community when compared to institutions although presumably survival would depend on the quality of the care package rather than the actual setting.4 Overall there is a wide range of other factors that may come into play that may affect life expectancy in any given individual.

It can be seen from this brief discussion that rehabilitation clearly has a role to play not only in the improvement of quality of life but also in terms of prolonging life expectation. The impact of many of the factors outlined above can be reduced by good quality multidisciplinary rehabilitation and by long term monitoring by a rehabilitation team. There is certainly evidence of improving survival in recent years. It is particularly obvious in the context of spinal cord injury. De Vivo and colleagues5, for example, showed that from 1973 to 1986 there was a 66% decrease in the risk of dying in the first two years post-injury. There have been similar improved trends of survival of severely disabled children6.

Whilst there is little argument about the key factors that reduce life expectancy there is considerable debate about quantifying life estimates. There are many variables. In addition to those outlined above other papers7have shown an improvement in life expectancy with regard to a variety other factors such as employment, community integration and a reasonable level of income. In an ideal world there would be large scale cohort studies that followed up significant numbers of disabled people for long enough for 50% to have died which would enable a median survival to be calculated.

Regrettably, such studies do not exist although survival trends can be implied from some of the larger scale publications. In the context of traumatic brain injury, for example, there are very few large scale cohort studies. Readers interested in this specific topic are referred to three up-to-date papers indicating a fairly modest reduction in life expectancy in those who have survived in the short term after traumatic brain injury.8910These papers compare to the somewhat more pessimistic results in the studies by Strauss and colleagues.11

There seem to be two schools of thought when an estimate of life expectancy is required. One is to use the periodic methodology. In this procedure a large number of disabled people are monitored over a short period of time (usually 1-3 years) and the numbers of deaths over this period are observed.

Thus, the risk of death over a particular age band can be calculated and a snapshot picture of the mortality of the whole population can be derived by various statistical methods.

The only sizeable databases in the world are those held by the Life Expectancy Project under the leadership of Professor David Strauss in California as well as the database on cerebral palsy held in the UK by Professor Jane Hutton and colleagues and the large database of spinal cord injury persons held in the USA by Michael De Vivo.

There are a number of problems in relying on such databases. There is an inbuilt pessimism bias as one cannot take into account future improvements in rehabilitation and nor do such “snapshot” databases take into account the more subjective factors involved in life expectancy, such as quality of care. An alternative methodology is to use existing country life tables and make a discount according to the various negative factors.

The disadvantage of this method is, obviously, that it depends very heavily on clinical interpretation of the importance of individual factors but it does allow a range of factors to be taken into account. Whether one uses the periodic methodology (bottom up) or the life table methodology (top down) approach then there is no doubt that at the present state of our knowledge clinical interpretation of the factors important for a given individual is still important. Whilst we await more accurate statistics the art of medicine in this arena is still paramount.

Hutton JL, Pharoah POD. Effects of cognitive, motor, and sensory disabilities on survival in cerebral palsy. Arch Dis Child 2002; 86: 84-90

De Vivo MJ, Stover SL, Black KJ. Prognostic factors for 12-year survival after spinal cord injury. Arch Phys Med Rehabil 1992; 73: 156-162

Frankel HL, Coll JR, Charilfue SW, et al. Long-term survival in spinal cord injury: a fifty year investigation.Spinal Cord 1998; 36: 266-274

Shavelle R, Strauss D, Day S. De-institutionalisation in California: mortality of persons with development disabilities after transfer into community care, 1997-1999. Journal of Data Science 2005; 3: 371-380

De Vivo MJ, Rutt RD, Black KJ, et al. Trends in spinal cord injury demographics and treatment outcomes between 1973 and 1986. Arch Phys Med Rehabil 1992; 73: 424-430

Strauss D, Shavelle R, Reynolds R et al. Survival in cerebral palsy in the last 20 years: signs of improvement?Dev Med Child Neurol 2007; 49: 86-92

Krause JS, De Vivo MJ, Jackson AB. Health status, community integration and economic risk factors for mortality after spinal cord injury. Arch Phys Med Rehabil 2004; 85: 1764-1773

Brown AW, Leibson CL, Malec JF et al. Long-term survival after traumatic brain injury: a population-based analysis. Neurorehabilitation 2004; 19(1): 37-43

Pentland B, Hutton LS, Jones PA. Late mortality after head injury. J Neurol Neurosurg Psychiatry 2005; 76(3): 395-400

Harrison-Felix C, Whiteneck G, De Vivo M, et al. Mortality following rehabilitation in the Traumatic Brain Injury Model Systems of Care. Neurorehabilitation 2004; 19(1): 45-54

Shavelle RM, Strauss DJ, Day SM, Ojdana KA. Life expectancy: In: Brain Injury Medicine: Principles and Practice. Zasler ND, Katz DI & Zafonte RD (Eds). Demos, New York 2007; 247-261