Degrees to Dollars: Health Costs Associated with Suboptimal Ambient Temperature Exposure

Suboptimal ambient temperature exposure is a public health issue affecting a large number of individuals globally. Given the influence of climate change, temperature is predicted to be more variable and extreme in the future, increasing concerns for relevant health risks. Although substantial research has focused on risk assessment, few studies have examined population health outcomes from an economic perspective. To inform climate-related health policies, we estimated the annual economic costs of adverse health outcomes associated with suboptimal ambient temperature exposures in the Minneapolis/St. Paul Twin Cities Metropolitan Area (TCMA). This area is the second largest economic center in the Midwestern region of the United States. Moreover, the TCMA is located in a continental climate experiencing a broad range of annual temperatures. To meet different policy needs, both extreme temperatures (defined as the bottom and top 5% temperature distribution) and moderate to extreme temperatures (defined as the bottom and top 30% temperature distribution) were considered.

We used a distributed lag nonlinear model combined with a Poisson model that allows for over-dispersion to derive the exposure-response functions between temperature levels and different measures of population health (i.e., mortality, emergency department visits, and emergency hospitalizations). This model further adjusted for long-term trends, weekday effects, and public holiday effects. Stratified analyses by age group (youth: 0-19 years, adults: 20-64 years, and elderly: 65+ years) were also included. Using these exposure-response functions, we calculated the numbers of mortality and morbidity associated with different magnitudes of suboptimal ambient temperature exposure. Then, we developed a comprehensive cost function to estimate the total economic costs of health outcomes related to suboptimal ambient temperature exposure from a societal perspective. Both direct and indirect costs were considered.

Exposure-response functions by population health measure and by age group indicate that the statistically significant association between suboptimal ambient temperature and mortality is driven by the mortality among the elderly. Extreme heat exposure only affects morbidity among youth. However, moderate and extreme cold exposures affect all age groups, although with different magnitudes. Relative risks increase most rapidly for youth as exposure approaches the lowest extreme temperature. Using these exposure-response functions, we discovered that youth have a relatively larger number of temperature-related emergency department visits, and the elderly have a relatively larger number of temperature-related mortality and emergency hospitalizations.

From an economic perspective, exposure to extremely low and high temperatures (the bottom and top 5% temperature distribution) leads to $2.70 billion [95% empirical confidence interval (eCI): $1.91 billion, $3.48 billion] ($2016) in economic costs annually. Of this total amount, $2.03 billion [95% eCI: $1.31 billion, $2.73 billion] is associated with extreme cold exposure and $0.67 billion [95% eCI: $0.34 billion, $0.99 billion] is associated with extreme heat exposure. Moderate-to-extreme low and high temperature exposures (the bottom and top 30% temperature distribution) bring the total estimate up to $9.40 billion [95% eCI: $6.05 billion, $12.57 billion] ($2016). Of this amount, $8.21 billion [95% eCI: $4.91 billion, $11.36 billion] is associated with moderate-to-extreme cold exposure and $1.17 billion [95% eCI: $0.61 billion, $1.75 billion] is associated with moderate-to-extreme heat exposure. Moreover, the majority of the economic costs are linked to mortality, as opposed to morbidity. Although the largest number of individuals affected by suboptimal ambient temperature exposure are 0-19 years old, elderly (65+ years) are the biggest contributor to the overall economic costs due to their heightened risk of mortality.

This study has immediate policy implications regarding the strategic prioritization of different age groups in intervention programs (e.g., risk communication and education). Specific recommendations will depend on the objectives of the decision maker. For instance, targeting youth is justifiable when the goal is to protect the most affected populations. Targeting seniors, especially when exposed to cold, may be more efficient in reducing the overall economic costs.

Beyond policy implications, the value of this study is three major contributions. First, it evaluates suboptimal ambient temperature exposure as a continuous variable, instead of discrete extreme exposure events (e.g. heatwaves or cold snaps). It shows that in addition to extreme temperatures, moderate temperature exposures can also lead to substantial health-related economic costs. Second, it adopts a multi-criteria approach to estimate the health-related economic costs of suboptimal ambient temperature exposures. The World Health Organization recommends this approach as a means of internalizing an array of external costs, enabling comparison across different outcomes, and providing explicit rules for balancing a range of information. Third, it develops a transparent and flexible cost function that is easily applicable to studying ambient temperature exposure in other populations and climates. When new parameters become available, this cost function can be easily updated. This framework can also be applied to assessing the impact of other environmental exposures that involve multiple health outcomes and economic costs, such as ambient air pollution.