External Causes of Death in Scotland: How Much Do They Contribute to Increasing Variation in Age-at-Death?

Rosie Seaman, Max Planck Institute for Demographic Research

Mortality outcomes in Scotland are poor within the context of Western Europe. Following the 1980s it experienced increasing premature mortality rates from suicide, substance and alcohol abuse, and accidents. These unfavourable mortality patterns are reflected in its life expectancy trend which has demonstrated the slowest relative rate of improvement. An additional, but lesser studied, dimension of mortality is variation in age-at-death. Higher variation in age-at-death is interpreted as higher inequality. Lower variation in age-at-death means age at death is more homogenous and is beneficial in terms of planning savings, pensions and health care needs. In a study of 17 Western European countries Scotland demonstrated the longest, sustained stagnating trend in variation in age-at-death. It is hypothesized that part of the explanation for Scotland’s trend is persisting premature adult mortality rates from external causes of death. This is the first study to estimate the age and cause-specific contributions to Scotland’s trend using individual level mortality data, covering the whole population from 1979 to 2016.

Theme:Mortality and longevity

External causes of death in Scotland:how much do they contribute to increasing variation in age-at-death?


All-cause mortality has improved for Scotland but some prematuremortality rates from suicide, substance and alcohol abuse, and accidents increasedfollowing the 1980s.1,2 This mortality pattern is reflected inScotland’s life expectancy which is now the lowest in Western Europe.3,4Life expectancy is one of the most common indicators of population health.5Increasing life expectancy means that the average age-at-death is beingdelayed. A lesser studied but equally important dimension of population level mortalityto measure over time is lifespan variation.6,7 Lifespanvariation estimates the amount of inequality in age-at-death that existsbetween individuals. Decreasing lifespan variation means that deaths are beingcompressed around a common age, and that age-at-death is becoming morehomogenous.6,8 We are the first to formally analyze the age and cause-specificcontributions to Scotland’s lifespan variation trend using the most robust andup-to-date mortality data available. Identifying theages and causes of death driving national trends can identify where mortalityreductions need to be achieved to reduce inequality and improve averagepopulation health.

Data and methods

The Human Mortality Database provides annualpopulation estimates for Scotland. Individual level mortality data covering theentire population of Scotland have been obtained via a commissioned requestfrom National Records of Scotland (1979-2016). Five broad cause of deathcategories (external causes, cancers, circulatory diseases, respiratorydiseases and other) will be constructed from harmonized ICD-9 and ICD-10 codes.Age-cause specific decomposition analysis will be performed using theadapted version of the Visual Basics Application program provided by Shkolnikovand Andreev.9

Preliminary results

Table 1 shows the change in lifespanvariation for 5-year time periods starting in 1970, for males and females inScotland (results produced using the all-cause mortality data available fromthe HMD). At the earliest time points, males experienced a decreasing trend. Atthe later time points, the rate of reduction slows and between 1990-94 and2005-09 diverges to slightly increasing for males. There is indication that lifespanvariation for males may once again be decreasing. For females the pattern isless clear. The data obtained for further analysis will allow us to report themost up-to-date trend and account for changes in the age and cause-specificmortality rates.

Future research direction

More detailed crossnational comparative research is needed to understand if the same ages andexternal causes of death driving Scotland’s national trend are similar incomparable populations. It may be that the substances most likely to causepremature deaths vary between countries because of differences in social,political, economic or healthcare contexts. It is our intention to use theHuman Cause of Death database (HCD) to facilitate future cross nationalcomparative research.  


Increasing lifespan variation, asidentified for males in Scotland, represents increasing inequality in age-of-deathat the population level. This trend may be the consequence of persistingpremature mortality rates. Cause of death analysis will be included in thefinal paper to increase our understanding of why population health in Scotlandis relatively poor compared to the rest of Western Europe (data obtained andanalysis in progress). Reducing premature mortality rates would improveScotland’s overall life expectancy and reduce lifespan variation which reflectsthe twin aim of public health: to simultaneously improve average health andreduce inequalities


1.           Leyland, A.H., Increasing inequalities in prematuremortality in Great Britain. Journal of Epidemiology and Community Health,2004. 58(4): p. 296-302.

2.           Schofield, L., et al., Dying younger in Scotland:Trends in mortality and deprivation relative to England and Wales, 1981–2011. Health& Place, 2016. 40: p. 106-115.

3.           McCartney, G., et al., Has Scotland always been the‘sick man’ of Europe? An observational study from 1855 to 2006. The EuropeanJournal of Public Health, 2012. 22(6): p. 756-760.

4.           Leyland, et al., Cause-specific inequalities inmortality in Scotland: two decades of change. A population-based study. BMCPublic Health, 2007. 7(1): p. 172-184.

5.           World Health Organization, World health statistics20102010: World Health Organization.

6.           Smits, J. and C. Monden, Length of life inequalityaround the globe. Social Science & Medicine, 2009. 68(6): p.1114-1123.

7.           Shkolnikov, et al., Losses of expected lifetime in theUnited States and other developed countries: methods and empirical analyses. Demography,2011. 48(1): p. 211-239.

8.           Tuljapurkar, S., The final inequality: variance inage at death, in Demography and the Economy. 2010, University of ChicagoPress. p. 209-221.

9.           Shkolnikov and E.M. Andreev, Age-decomposition of adifference between two populations for any life-table quantity in Excel, MaxPlanck Institute for Demographic Research., Editor 2010, : Rostock

Presented in Session 1233: Mortality and Longevity