non-disponible en francais

VitaMins Health: Getting clued-up on the flu

VitaMins Health: Getting clued-up on the flu

October 2019

Every year significantly more people die during the winter months than at other times of the year, and we have recently seen huge variability in ‘excess winter deaths’, potentially impacted by the severity of the flu.

With annual volatility in death rates being largely driven by the level of excess winter deaths, and potentially more than a third caused by respiratory diseases1, can we use an understanding of influenza patterns to help predict future events? And, how much should annual fluctuations of data due to the effectiveness of a vaccine for the most prevalent flu strain affect long-term projections?

What is seasonal influenza?

Influenza is a virus that affects the respiratory system. It is usually not life-threatening, however, certain groups – such as those over age 65 and under age 5, pregnant women and the severely obese – are far more at risk of life-threatening complications. Over the five years ending in 2017, 90% of Canadian influenza deaths were for those age 65 and older².

Most influenza infections occur during the winter months due to a combination of factors, potentially including:

  • A decline in ultraviolet (UV) radiation, reducing the chances of the virus being killed by UV rays.
  • Decreased production of Vitamin D, vital for the immune system, due to decreased exposure to sunlight.
  • People staying indoors more, increasing the likelihood of transmission.

The two most common types of the influenza virus that infect humans are known as influenza A and B, with influenza A being more common and generally more severe. There are also many different subtypes of the virus, with the two most prevalent subtypes of influenza A being H1N1 and H3N2.

How does flu affect mortality?

 Deaths by season for Canada

In the chart above shows the number of deaths in Canada, by season, for 2010 to 2017 (the latest year currently available)3.

The purple bars represent winter deaths, showing a spike in deaths from December to February each year (i.e., 2017 means December 2016 to February 2017).

Using data from the World Health Organization (WHO), we have overlaid which flu subtype/type (i.e., H1N1, H3N2 or B) was the most prevalent each year (by number of identified cases). The chart illustrates that in years where H3N2 has been the dominant flu strain, the level of excess winter deaths has been higher, suggesting that H3N2 has recently been more life-threatening than H1N1. However, continual mutations of the virus mean this effect may well change over time.

Another way of measuring excess winter deaths is by the ratio of winter to summer deaths, with the bars in the following chart showing the ratio of winter to summer deaths in Canada for the past 15 years3. The chart highlights that the annual volatility of excess winter deaths in Canada has increased since 2010, and 2015 was especially high. The pink line in the chart plots the annual number of Canadian deaths attributable to influenza and pneumonia2 – we can see that the number of influenza and pneumonia deaths is closely aligned with the ratio of winter to summer deaths.

Top charts

The flu vaccine: choice of strains and the production process

Each winter, various public health programs promote vaccination against influenza for at-risk groups. The vaccine is designed to combat the most prevalent strain of the virus that year. Due to rapid mutations of the influenza virus, the vaccine’s effectiveness will vary and likely decrease over time.

So, how are vaccine strains chosen? Organizations around the world, such as the Public Health Agency of Canada, the Centers for Disease Control and Prevention (CDC) in the US and Public Health England in the UK, collect data on circulating influenza strains and monitor the virus’ genome and structure. In February each year, the WHO decides which strains are most likely to be prevalent in the Northern Hemisphere next winter, based on information such as the strains currently prevalent in the Southern Hemisphere and those which have dominated in recent years (there is a similar process for the Southern Hemisphere in August). However, this estimate is not always correct; the high number of excess winter deaths in 2014 / 2015 was widely attributed to a mis-match in the vaccine and the most prevalent flu strain4.

The vaccine strains are chosen in February for the following winter because producing large quantities of influenza vaccine takes at least six months. The vaccine is typically made inside fertilized chicken eggs, one egg per vaccine dose, which is a very slow process.

The future of the flu

The impact of the flu on mortality will be largely dependent on the ability to effectively vaccinate against it, not to mention the willingness of people to be vaccinated. The Public Health Agency of Canada estimated flu vaccine coverage among Canadian adults to only be 38% for the 2017 / 2018 season, however the coverage rate was 71% for those age 65 and older5. Achieving an effective flu vaccine has many challenges, which various advances in treatment are trying to combat, and may in turn increase vaccination coverage rates.

Universal vaccine
A universal vaccine is one that would protect an individual from all influenza viruses. There is currently no universal flu vaccine licensed for human use. However, there are some promising drugs in clinical trials6, although these are still at least five years away from commercial use.

Vaccine production
There are two newer production processes7 for creating the influenza vaccine: a cell-based method and a form of DNA manufacturing known as recombinant. Both methods are faster than the egg-based process, with recombinant being the fastest. Adopting these faster processes would allow the WHO to choose the influenza vaccine strains later, with a more accurate picture of the prevalent strains. This would increase vaccine efficacy and ultimately save lives.

Despite recombinant technology offering faster production, there remain some issues8 preventing its wide adoption, particularly: cost, a short shelf-life and insufficient infrastructure to mass-produce globally

What does this mean for pensioner longevity?

The number of deaths experienced each year drives our best estimates for how long people will live both now and in the future. The relative number of deaths experienced in any particular year can largely be affected by the level of excess winter deaths, which in turn is driven by the most prevalent flu strain and the efficacy of that year’s vaccine. Monitoring common flu strains in other parts of the world and advances in vaccine effectiveness could give early warnings to data effects that will feed into future models.

We should also be wary of models that are too sensitive to a single year’s experience data. We should make sure to consider whether a mis-match between a particular year’s most prevalent influenza strain and the strain covered by that year’s vaccine may be inappropriately affecting our long-term outlook on longevity improvements

1Excess winter mortality in England and Whales: 2016 to 2017 (provisional) and 2015 to 2016 (final), November 22, 2017 (link)

2Statistics Canada. Table 13-10-0782-01 Deaths, by cause, Chapter X: Diseases of the respiratory system (J00 to J99)

3Statistics Canada. Table 13-10-0708-01 Deaths, by month

4Flu vaccine no match against bug that popped up near end, June 27, 2019,

5Seasonal Influenza Vaccine Coverage in Canada, 2017-2018, January 2019.

6Universal Influenza Vaccine Research, National Institute of Allergy and Infectious Diseases,

7Influenza Vaccine Production and Design, National Institute of Allergy and Infectious Diseases,

8Efforts to Improve the Seasonal Influenza Vaccine, March 30, 2018,