• Research and White Papers
  • September 2019
  • 5 minutes

Update: Unnatural Cause of Death Experience Study

An in-depth analysis of the World Health Organization Mortality Database

  • Peter Banthorpe
  • Taylor Pickett
  • Jason McKinley
  • Jake Davis
Skip to Authors and Experts
View of pier facing lake at sunset
In Brief

RGA updated a 2015 analysis of World Health Organization (WHO) mortality data to develop a thorough analysis of causes of unnatural deaths across seven countries. The results revealed compelling trends across multiple causes, from motor vehicle accidents to suicides, homicides, and accidental poisoning/drug overdose.

We analyzed trends over a period of more than 35 years. The table below summarizes our key findings by cause of unnatural death:



Motor Vehicle Accidents

After years of significant improvement in all countries, the last few years show that improvement has leveled off or declined in most countries

Accidental Poisoning (incl. drug overdoses)

Significant worsening in the United States (U.S.), United Kingdom (U.K.), Canada, and France. France is notable in that older people contributed more to the decline than younger people.


Low rates in the U.K. and Italy. Highest in the U.S. and France. Suicide rates have steadily increased in the U.S. for the last 15 years, while steadily decreasing in France over the past 30 years. Canada has also seen rates increase in recent years.


Homicide rates are a problem limited mostly to the U.S. After years of general improvement, rates have begun climbing again in the most recent years. The increase in the U.S. between 2013 and 2016 alone would be larger than the total rate for most countries in the most recent year of data.

The 2015 report stated that the U.S. is the only country to see periods of unnatural death mortality deterioration. That is no longer the case. Canada, the U.K., and even Hong Kong SAR have experienced deteriorating unnatural death mortality since 2011.

This report provides a dedicated section on the opioid crisis, which has been widely discussed and researched in recent years, as well as details about other important trends related to unnatural causes of death.

Given the depth, breadth, and quality of the data in the WHO Mortality Database, we believe it will prove useful in many ongoing and future research endeavors. However, the user must be mindful of differences in how cause of death data is recorded among different countries and over time.


The World Health Organization (WHO) mortality database contains a time-series of mortality rates, classified by cause of death, for a large number of countries (World Health Organization, 2019). We used this database to calculate unnatural death rates for select countries with a combination of criteria:

  • Populations large enough to provide relatively stable results over time

  • Length of time-series available, to enable assessment of trends

We examined seven countries for our detailed report analysis. The table below shows the seven countries selected and our rationale for each. The table is ordered broadly with the countries experiencing the highest unnatural death rates at the top of the table:



Other Interesting Features


Highest unnatural death rates among large developed countries.

Despite high rates, France has experienced consistent improvement in the last 30 years.

United States

High levels of unnatural death rates were followed by a period of improvement, which has deteriorated over the last 15 years.

Accidental Poisoning rates have deteriorated.

Significantly higher level of homicide deaths compared to other countries.




Moderate experience initially, followed by relatively strong improvement.


Hong Kong

Consistently among the very lowest rates in the past 30 years. Another option here would have been the Netherlands.




After an initial deterioration of rates, Singapore's rates have improved relative to that of Hong Kong SAR and the U.K.



Explore the report to examine all seven countries in greater detail. For France, Canada, Italy, the U.K., and the U.S., we have sufficient data to extend this analysis down to an age-banded analysis. All results include both sexes.

More Like This...

Meet the Authors & Experts

Peter Banthorpe
Peter Banthorpe
Managing Director, RGA UK
Taylor Pickett
Taylor Pickett
Actuary, Pricing, U.S. Individual Life, RGA
Jason McKinley
Jason McKinley
Actuary, Risk and Behavioral Science
Jake Davis
Global Data and Analytics, RGA