Trends in Excess Winter Mortality (EWM) from 1900/01 to 2019/20-Evidence for a Complex System of Multiple Long-Term Trends.

Rodney P. Jones,Andriy Ponomarenko

doi:10.3390/ijerph19063407

Abstract

Trends in excess winter mortality (EWM) were investigated from the winter of 1900/01 to 2019/20. During the 1918–1919 Spanish flu epidemic a maximum EWM of 100% was observed in both Denmark and the USA, and 131% in Sweden. During the Spanish flu epidemic in the USA 70% of excess winter deaths were coded to influenza. EWM steadily declined from the Spanish flu peak to a minimum around the 1960s to 1980s. This decline was accompanied by a shift in deaths away from the winter and spring, and the EWM calculation shifted from a maximum around April to June in the early 1900s to around March since the late 1960s. EWM has a good correlation with the number of estimated influenza deaths, but in this context influenza pandemics after the Spanish flu only had an EWM equivalent to that for seasonal influenza. This was confirmed for a large sample of world countries for the three pandemics occurring after 1960. Using data from 1980 onward the effect of influenza vaccination on EWM were examined using a large international dataset. No effect of increasing influenza vaccination could be discerned; however, there are multiple competing forces influencing EWM which will obscure any underlying trend, e.g., increasing age at death, multimorbidity, dementia, polypharmacy, diabetes, and obesity—all of which either interfere with vaccine effectiveness or are risk factors for influenza death. After adjusting the trend in EWM in the USA influenza vaccination can be seen to be masking higher winter deaths among a high morbidity US population. Adjusting for the effect of increasing obesity counteracted some of the observed increase in EWM seen in the USA. Winter deaths are clearly the outcome of a complex system of competing long-term trends.

Highlights

A previous study in IJERPH investigated the use of excess winter mortality (EWM) as a forensic tool [1]
While the EWM calculation for all three countries during the Spanish flu shows a maximum for the 4 months ending January 1919, the USA and Sweden shows maximum monthly deaths in October 1918, while maximum deaths in Denmark occurred in November 1918
Given the fact that the USA has among the highest levels of obesity, diabetes, and cancer in the developed world we proposed that EWM may be trending upward over time in the US

Summary

Introduction

A previous study in IJERPH investigated the use of excess winter mortality (EWM) as a forensic tool [1]. The magnitude of EWM was shown to be greatly influenced by respiratory deaths and by a mix of other factors such as access to health care, home insulation (indoor temperature), and age at death [1]. All these factors are changing over time. There has been considerable debate around exactly how many influenza deaths have occurred during various pandemics. The difficulty is that such deaths must be estimated and that different estimation methods give widely different answers [2–8]. Deaths due to influenza are a subset of EWM, and the high dependence of EWM on respiratory deaths [1]

Methods

Results

Discussion

Conclusion