Objective Gains in life expectancy have faltered in several high-income countries in recent years. We aim to compare life expectancy trends in Scotland to those seen internationally, and to assess the timing of any recent changes in mortality trends for Scotland. Setting Austria, Croatia, Czech Republic, Denmark, England & Wales, Estonia, France, Germany, Hungary, Iceland, Israel, Japan, Korea, Latvia, Lithuania, Netherlands, Northern Ireland, Poland, Scotland, Slovakia, Spain, Sweden, Switzerland, USA. Methods We used life expectancy data from the Human Mortality Database (HMD) to calculate the mean annual life expectancy change for 24 high-income countries over five-year periods from 1992 to 2016, and the change for Scotland for five-year periods from 1857 to 2016. One- and two-break segmented regression models were applied to mortality data from National Records of Scotland (NRS) to identify turning points in age-standardised mortality trends between 1990 and 2018. Results In 2012-2016 life expectancies in Scotland increased by 2.5 weeks/year for females and 4.5 weeks/year for males, the smallest gains of any period since the early 1970s. The improvements in life expectancy in 2012-2016 were smallest among females (<2.0 weeks/year) in Northern Ireland, Iceland, England & Wales and the USA and among males (<5.0 weeks/year) in Iceland, USA, England & Wales and Scotland. Japan, Korea, and countries of Eastern Europe have seen substantial gains in the same period. The best estimate of when mortality rates changed to a slower rate of improvement in Scotland was the year to 2012 Q4 for males and the year to 2014 Q2 for females. Conclusion Life expectancy improvement has stalled across many, but not all, high income countries. The recent change in the mortality trend in Scotland occurred within the period 2012-2014. Further research is required to understand these trends, but governments must also take timely action on plausible contributors.

Description of methods used for collection/generation of data:
The HMD has a detailed methods protocol available here:
https://www.mortality.org/Public/Docs/MethodsProtocol.pdf
The ONS and NRS also have similar methods for ensuring data consistency and quality assurance.

Methods for processing the data:
The segmented regression was conducted using the 'segmented' package in R. The recommended references to this package and its approach are here:
  Vito M. R. Muggeo (2003). Estimating regression models with unknown break-points. Statistics in Medicine, 22, 3055-3071.

  Vito M. R. Muggeo (2008). segmented: an R Package to Fit Regression Models with Broken-Line Relationships. R News, 8/1,
  20-25. URL https://cran.r-project.org/doc/Rnews/.

  Vito M. R. Muggeo (2016). Testing with a nuisance parameter present only under the alternative: a score-based approach with
  application to segmented modelling. J of Statistical Computation and Simulation, 86, 3059-3067.

  Vito M. R. Muggeo (2017). Interval estimation for the breakpoint in segmented regression: a smoothed score-based approach.
  Australian & New Zealand Journal of Statistics, 59, 311-322.

Software- or Instrument-specific information needed to interpret the data, including software and hardware version numbers:
The analyses were conducted in R version 3.6.1 and Microsoft Excel 2013.

Please see README.txt for further information