Skip to main content
White Paper

Climate variables as drivers of mortality

The mortality impacts of climate change - part 2 of a 5-part series

Climate Quantified|Insurance Consulting and Technology
Climate Risk and Resilience|Climate and Resilience Hub|Insurer Solutions

By Richard Marshall | March 4, 2021

This paper investigates the liability-side impact of climate change. The second chapter explores the concept of a driver of mortality and how this can be used in the context of climate risk.

Unlock More

About this Series

'The mortality impacts of climate change' investigates the liability-side impact of climate change, focusing on the effects on UK life insurers. It considers the mechanisms through which climate can affect mortality, models and quantifies some of those effects and discusses the implications for the life insurance industry.

What do we mean by a driver of mortality?

A ‘driver of mortality’ can be understood to be a variable with a material link to mortality rates. Commonly used drivers of mortality in the insurance world include smoking prevalence, diet and exercise habits, pharmaceutical developments and public policy relating to health and social care spending. However, in recent years there has been a marked increase in research activity directed at better understanding of the links between climate variables – such as temperature, precipitation or extreme weather – and health outcomes (World Health Organisation, 2018).

Long-listing climate variables and the evidence for a link between climate and mortality

Climate affects a range of geophysical, ecological and socio-economic systems which influence human health. Some effects may be relatively simple and easy to quantify, such as increased mortality from heatwaves, but others, particularly indirect impacts such as changes in migration patterns or food security, may be complex and difficult to model. It is also important to note that climate impacts do not occur in isolation but in the context of other large-scale societal and environmental factors such as changes in land use, biodiversity, urbanisation and economic growth or decline. These factors may also affect patterns of health and disease, both directly and via their influence on climate and society’s responses (Department of Health, 2001). This chapter provides a brief overview of current evidence for a link between long-listed climate variables and mortality.

Extreme temperature

An extensive body of literature documents the relationship between extreme temperatures (usually during heat waves, but also cold waves) and mortality (Department of Health, 2001; Baccini, et al., 2008; Ishigami, et al., 2008; Hajat & Kosatky, 2010). The temperature–mortality relationship is usually a nonlinear U-, V-, or J-shape. Many studies have quantified cold and heat effects separately, assuming a linear response below and above a threshold temperature (Baccini, et al., 2008; McMichael, et al., 2008; Hajat & Kosatky, 2010). The health effects of heat can be estimated using the heat threshold (the temperature at which the harmful effect of heat begins to occur), and the heat slope (a measure of the size of this effect) (Hajat & Kosatky, 2010)). A significant geographic variability has been observed in both heat thresholds and slopes. Heat thresholds tend to be higher in warmer locations, suggesting acclimatisation (Baccini, et al., 2008; Medina-Ramon & Schwartz, 2007).

Exposure to extreme heat has been associated with deaths from cardiovascular, respiratory, and cerebrovascular diseases, particularly among elderly people (Stafoggia, 2006; Barnett, 2007; Ishigami, et al., 2008; McMichael, et al., 2008; Hertel, et al., 2009). This is because the need for body temperature regulation at high temperatures imposes additional stress on the cardiovascular and respiratory systems (Kovats & Hajat, 2008; WHO, 2009). There is also evidence of a link between high temperatures and non-medical deaths (i.e. accidents, suicides), however the exact mechanism by which heat affects mortality from these causes is unclear and needs further investigation (Page, et al., 2007).

Cold is also a risk factor for mortality, although impacts are generally delayed. Exposure to very cold temperatures causes cardiovascular stress due to changes in blood pressure, vasoconstriction, and an increase in blood viscosity (which can lead to clots), as well as higher levels of red blood cell counts, plasma cholesterol, and plasma fibrinogen (Huynen et al. 2001). Further, susceptibility to pulmonary infections may increase because breathing cold air can lead to bronchoconstriction. Unlike hot days, however, health conditions associated with extreme cold have been shown to take longer to manifest themselves and spread, meaning that the full effect of a cold day takes a few weeks to manifest itself (Analitis, et al., 2008).

Air pollution

There is a large body of research that definitively demonstrates the link between air pollution – defined here as any chemical, physical (particulate matter), or biological agent that modifies the natural characteristics of the atmosphere – and mortality (Daniels, et al., 2000; Næss, et al., 2007; Bowe, et al., 2019). It is associated with an increase in cardiovascular diseases, such as heart attacks and strokes, as well as aggravated asthma, bronchitis, emphysema, lung diseases, and respiratory allergies.

Particulate matter (PM) and ozone are pollutants of special concern. Bowe, et al., (2019) showed a direct relationship between exposure to PM and negative health impacts. Smaller-diameter particles (PM2.5 or smaller) are generally more dangerous and ultrafine particles (one micron in diameter or less) can penetrate tissues and organs, posing an even greater risk of systemic health impacts.

Similarly, many studies have shown that ozone can cause the muscles in the airways to constrict, trapping air in the alveoli (EPA, 2020). In the short-term, this can aggravate lung diseases such as asthma, emphysema, and chronic bronchitis, make the lungs more susceptible to infection and cause chronic obstructive pulmonary disease (COPD). Long-term exposure is linked to aggravation of asthma and is likely to be one of many causes of asthma development. It may also be linked to permanent lung damage, such as abnormal lung development in children, but the evidence is not as strong as the evidence for short-term exposure (Bowe, et al., 2019).


There is a growing body of research documenting a link between windstorms and cause-specific mortality, although the evidence is not as strong as the evidence for extreme temperature or air pollution. Goldman, et al., (2014) provide a comprehensive overview of this link, finding that direct effects occur during the impact phase of a storm, causing death and injury due to the force of the wind, with becoming airborne, being struck by flying debris or falling trees and road traffic accidents being the main dangers. Indirect effects, occurring during the pre- and post-impact phases of the storm, include falls, lacerations and puncture wounds, and occur when preparing for, or cleaning up after a storm. Power outages are a key issue and can lead to electrocution, fires and burns and carbon monoxide poisoning from diesel or petrol powered electrical generators. Additionally, worsening of chronic illnesses due to lack of access to medical care or medication can occur.


Several studies have indicated that rainy conditions can have a significant impact on mortality risk from accidental causes, mainly road traffic accidents. Eisenberg & Warner (2005), for example, find that precipitation correlates with markedly increased crash rates. Recent work also shows that the risk posed by precipitation rises dramatically with the time since last precipitation. There is limited evidence of a direct relationship between precipitation and other causes of death in a UK context. However, intense rainfall can cause flooding which carries its own set of risk and which is discussed in the next sub-section.

The effects of a specific form of precipitation, snowfall, on mortality has also been explored, although only a handful of published studies, that have produced some conflicting results, exist. Eisenberg & Warner (2005) summarise these, noting that significantly increased crash rates have been documented in snowy months in Canada (Andreescu & Frost, 1998), on snowy days in the United Kingdom (Perry & Symons, 1991), and during snowstorms in Iowa (Knapp, et al., 2000). Perry and Symons found increased rates of crashes involving injuries and fatalities on snowy days in the United Kingdom, but Brown and Baass (1997) noted fewer crashes involving injuries in the winter months in Canada. Eisenberg (2004) found decreased rates of fatal crashes on snowy days in the United States, a finding echoed in analysis of winter months in Canada and snowy months in Scandinavia (Brown & Baass, 1997).

A possible reason for this conflict is explained by Eisenberg & Warner (2005). On the one hand, they point out, snow makes driving more dangerous, by reducing tyre grip and impairing visibility, thereby increasing the risk of accident and possibly death. On the other hand, drivers typically drive more slowly and carefully in snowy weather, and many people avoid or postpone unnecessary travel, reducing the risk of a fatal accident. Overall, the weight of the evidence suggests that less severe crashes (e.g., those producing only property damage) increase during snow days, while more severe crashes (those resulting in fatalities) decrease.

At the opposite end of precipitation extremes, drought also poses risks to morbidity and mortality. The Center for Disease Control and Prevention notes that drought conditions may increase the environmental exposure to a broad set of health hazards including wildfires, dust storms, extreme heat events, flash flooding, degraded water quality, and reduced water quantity. Dust storms associated with drought conditions contribute to degraded air quality due to particulates and have been associated with increased incidence of coccidioidomycosis (valley fever), a fungal pathogen, in Arizona and California (CDC, 2020).


As noted above, extreme precipitation can contribute to severe flooding events which are a risk factor for mortality. Direct exposure to flood water can cause drownings and physical injuries, destroy homes, medical facilities and other essential services and lead to worsening of chronic illnesses due to lack of access to medical care or medication. However, in most high-income countries such as the UK, there are usually few immediate deaths except during the most exceptional flood events (Milojevic, et al., 2011).

Floods can also contaminate freshwater supplies, heighten the risk of water-borne diseases, and create breeding grounds for disease-carrying insects such as mosquitoes (although other variables may affect these associations). Furthermore, people living in damp indoor environments have been found to experience increased prevalence of asthma and other upper respiratory tract symptoms, such as coughing and wheezing, as well as lower respiratory tract infections such as pneumonia, respiratory syncytial virus (RSV), and RSV pneumonia (CDC, 2020).

Vector- and water-borne diseases

A growing body of literature documents the link between climate, a subset of vector- and water-borne diseases and mortality.

Vector-borne diseases are infections transmitted by the bite of infected arthropod species, such as mosquitoes (West Nile virus, Dengue fever, Chikungunya fever and malaria), ticks (tick-borne encephalitis, Human Granulocytic Anaplasmosis), triatomine bugs, sandflies (Leishmaniasis), tsetse flies (African sleeping sickness) and blackflies.

Water-borne diseases, as the name suggests, include many different types of infections that are transmitted via water and include pathogens across a range of taxa (viruses, bacteria, protozoa, and helminths) that can cause an array of symptoms, including diarrhoea, fever and other flu-like symptoms, neurological disorders and liver damage.

The relationship between climate and disease transmission is complex. Arthropod vectors are cold-blooded (ectothermic) and thus especially sensitive to climatic factors. The European Centre for Disease Prevention and Control notes that weather influences survival and reproduction rates of vectors, in turn influencing habitat suitability, distribution and abundance; intensity and temporal pattern of vector activity (particularly biting rates) throughout the year; and rates of development, survival and reproduction of pathogens within vectors (ECDC, 2020). For water-borne diseases, Levy, et al., (2018) note that exposure to high temperatures can change pathogen survival, replication and virulence, heavy rainfall events can mobilize pathogens and compromise water and sanitation infrastructure, and drought can concentrate pathogens in limited water supplies.

However, the impacts of climate on vector- and water-borne diseases depend not simply on meteorological conditions, but also on the underlying social and ecological contexts – from water and sanitation infrastructure to local pathogen distribution to social capital – that influence a population’s exposure, sensitivity and adaptive capacity. This makes it very difficult to quantify the impact of change in a climate variable on mortality rates. (The potential impact on climate change on vector-borne and water-borne disease is discussed further in Chapter 4.)

Short-listing of climate variables for modelling

To model the effects of climate change using a driver-based approach, a set of climate-related drivers of mortality is required for which there is evidence of a potentially material link to certain cause-specific mortality rates and a convincing link to changes in the climate. Of the climate drivers set out above a short-list (see Table 1) was developed with such evidence available. The driver-based model also requires a breakdown of overall mortality by cause of death. These groups of causes of death were selected based on the granularity of the evidence for the link between the drivers and mortality identified in literature and are summarised in Table 2, along with the corresponding ICD-10 codes.

Table 1. Short-listed climate drivers selected for inclusion in the driver-based model and their definition
Climate driver Definition
Extreme heat % of year at 21 °C or above
The evidence allows us to determine an approximate relative risk of death on days which exceed 21 °C compared to that on days not reaching 21 °C . A useful way to estimate the effect of temperature change on annual deaths is to consider the proportion of the year during which the higher risk applies and how this might change over time.
Extreme cold % of year at 1 °C or below
Similarly to above, the evidence allows us to determine an approximate relative risk of death on days where temperatures fall below 1 °C compared to that on days not falling below 1 °C. A useful way to estimate the effect of temperature change on annual deaths is to consider the proportion of the year during which the higher risk applies and how this might change over time.
Rainfall % of year with daily rainfall of 10mm or greater
The evidence allows us to determine an approximate relative risk of death on days where rainfall exceeds 10mm compared to that on days not reaching 10mm. A useful way to estimate the effect of rainfall change on annual deaths is to consider the proportion of the year during which the higher risk applies and how this might change over time.
Snowfall % of year with at least 3cm snowfall
The evidence allows us to determine an approximate relative risk of death on days where snowfall exceeds 3cm compared to that on days not reaching 3cm. A useful way to estimate the effect of snowfall change on annual deaths is to consider the proportion of the year during which the higher risk applies and how this might change over time.
Storms % of year during which a named storm is over the UK
The evidence allows us to determine an approximate relative risk of death on days where a named storm occurs compared to that on days where a named storm does not occur. A useful way to estimate the effect of storm change on annual deaths is to consider the proportion of the year during which the higher risk applies and how this might change over time. A named storm is defined here in accordance with the UK Storm Centre, which names a storm when it is deemed to have a "substantial" impact on the UK or Ireland, usually when mean wind speeds in excess of 80 km/h (50 mph) or gusts over 130 km/h (80 mph) are reported.
Air pollution % of year during which PM10 is above 35µg/m3, or O3 is above 42.5 ppb
The evidence allows us to determine an approximate relative risk of death on days where a specific air pollutant exceeds a given threshold compared to that on days where the threshold is not reached. A useful way to estimate the effect of air change on annual deaths is to consider the proportion of the year during which the higher risk applies and how this might change over time.

Table 2. Cause of death groups included in the driver-based model
Cause of death group ICD-10 code(s)
Cancer C00-C99, D00-D48
Ischaemic heart disease / circulatory I00-I59, I70-I99
Cerebrovascular disease I60-I69
Respiratory disease J00-J08, J19-J99
Infectious diseases (excluding HIV) A00-A99, B00-B19, B25-B99, J09-J18, U80-U89
Other medical conditions All codes starting with: E to H, K to T. Along with: B20-B24, D49-D99, U00-U79 and U90-U99.
Non-medical causes All codes starting with: V to Y

Next time

The third chapter in our paper describes our modelling approach in detail. Driver-based modelling is introduced, along with details of how such a model has been calibrated to examine climate risk. The development of climate change scenarios is set out and the consistency of our mortality projections with existing mortality projections is discussed.

Upcoming chapters

  • Driver-based modelling
  • Impacts of climate change
  • Implications for life insurers and pension schemes


Analitis, A. et al., 2008. Temperature effects on mortality: Potential confounding by air pollution and possible interactions within the PHEWE project.. Epidemiology, 19(1), p. 214.
Andreescu, M. & Frost, D., 1998. Weather and traffic accidents in Montreal, Canada. Climate Res. , Volume 9, pp. 1-20.
Baccini, M. et al., 2008. Heat effects on mortality in 15 European cities.. Epidemiology, pp. 711-719.
Barnett, A., 2007. Temperature and cardiovascular deaths in the US elderly: changes over time.. Epidemiology, pp. 369-372.
Bowe, B., Xie, Y., Yan, Y. & Al-Aly, Z., 2019. Burden of cause-specific mortality associated with PM2. 5 air pollution in the United States.. JAMA network open, 2(11).
Brown, B. & Baass, K., 1997. Seasonal variation in frequencies and rates of highway accidents as a function of severity.. Transportation Res Rec, Volume 1581, p. 59–65.
CDC, 2020. Climate and Health: Precipitation Extremes. [Online] Available at: [Accessed 2020].
Daniels, M., Dominici, F., Samet, J. & Zeger, S., 2000. Estimating particulate matter-mortality dose-response curves and threshold levels: an analysis of daily time-series for the 20 largest US cities. American journal of epidemiology, 152(5), pp. 397-406.
Department of Health, 2001. Health Effects of Climate Change in the UK, s.l.: Department of Health.
ECDC, 2020. Vector-borne diseases. [Online] Available at: [Accessed 2020].
Eisenberg, D., 2004. The mixed effects of precipitation on traffic crashes. Accid Anal Prev, Volume 36, p. 637–647.
Eisenberg, D. & Warner, K., 2005. Effects of snowfalls on motor vehicle collisions, injuries, and fatalities.. American journal of public health, 95(1), pp. 120-124.
EPA, 2020. Health Effects of Ozone Pollution. [Online] Available at:,ozone%20can%20cause%20health%20effects. [Accessed 2020].
Goldman, A., Eggen, B., Golding, B. & Murray, V., 2014. The health impacts of windstorms: a systematic literature review.. Public health, 128(1), pp. 3-28.
Hajat, S. & Kosatky, T., 2010. Heat-related mortality: a review and exploration of heterogeneity.. Journal of Epidemiology & Community Health, 64(9), pp. 753-760.
Hertel, S., Le Tertre, A., Jöckel, K. & Hoffmann, B., 2009. Quantification of the heat wave effect on cause-specific mortality in Essen, Germany.. European journal of epidemiology, 24(8), pp. 407-414.
Huang, C. et al., 2011. Projecting future heat-related mortality under climate change scenarios: a systematic review.. Environmental health perspectives, 112(12), pp. 1681-1690.
Ishigami, A. et al., 2008. An ecological time-series study of heat-related mortality in three European cities.. Environmental Health, 7(1), p. 5.
Knapp, K., Smithson, L. & Khattak, A., 2000. The mobility and safety impacts of winter storm events in a freeway environment. Mid-Continent Transportation Res Symp, pp. 67-71.
Kovats, R. & Hajat, S., 2008. Heat stress and public health: a critical review.. Annu. Rev. Public Health, Volume 29, pp. 41-55.
Levy, K., Smith, S. & Carlton, E., 2018. Levy, K., Smith, S.M. and Carlton, E.J., Climate change impacts on waterborne diseases: moving toward designing interventions.. Current environmental health reports, 5(2), pp. 272-282.
McMichael, A. et al., 2008. International study of temperature, heat and urban mortality: the ‘ISOTHURM’project. International journal of epidemiology, 37(5), pp. 1121-1131.
Medina-Ramon, M. & Schwartz, J., 2007. Temperature, temperature extremes, and mortality: a study of acclimatisation and effect modification in 50 US cities.. Occupational and environmental medicine, 64(12), pp. 827-833.
Milojevic, A. et al., 2011. Long-term effects of flooding on mortality in England and Wales, 1994-2005: controlled interrupted time-series analysis.. Environmental Health, 10(1), p. 11.
Næss, Ø. et al., 2007. Relation between concentration of air pollution and cause-specific mortality: four-year exposures to nitrogen dioxide and particulate matter pollutants in 470 neighborhoods in Oslo, Norway.. American journal of epidemiology, 165(4), pp. 435-443.
Page, L., Hajat, S. & Kovats, R., 2007. Relationship between daily suicide counts and temperature in England and Wales.. The British Journal of Psychiatry, 191(2), pp. 106-112.
Perry, A. & Symons, L., 1991. Highway Meteorology. Swansea, Wales: E & F N Spon Press.
Stafoggia, M. et al., 2006. Vulnerability to heat-related mortality: a multicity, population-based, case-crossover analysis.. Epidemiology, pp. 315-323.
United Nations Office for Disaster Risk Reduction, 2017. UNISDR Strategic Framework 2016-2021. [Online] Available at: [Accessed 19 January 2021].
WHO, 2009. Improving Public Health Responses to Extreme Weather Heat-Waves EuroHEAT., s.l.: World .
World Health Organisation, 2018. Climate change and health. [Online] Available at:,malaria%2C%20diarrhoea%20and%20heat%20stress. [Accessed 27 October 2020].


Richard Marshall

Richard Marshall is a Director in Willis Towers Watson’s Insurance Consulting and Technology business and leads the development of mortality and demographic risk models for our UK business.

Related Solutions

Contact Us