Heat/mortality sensitivities in Los Angeles during winter: a unique phenomenon in the United States

Abstract

Background: Extreme heat is often associated with elevated levels of human mortality, particularly across the mid-latitudes. Los Angeles, CA exhibits a unique, highly variable winter climate, with brief periods of intense heat caused by downsloping winds commonly known as Santa Ana winds. The goal is to determine if Los Angeles County is susceptible to heat-related mortality during the winter season. This is the first study to specifically evaluate heat-related mortality during the winter for a U.S. city.

Methods: Utilizing the Spatial Synoptic Classification system in Los Angeles County from 1979 through 2010, we first relate daily human mortality to synoptic air mass type during the winter season (December, January, February) using Welch's t-tests. However, this methodology is only somewhat effective at controlling for important inter- and intra-annual trends in human mortality unrelated to heat such as influenza outbreaks. As a result, we use distributed lag nonlinear modeling (DLNM) to evaluate if the relative risk of human mortality increases during higher temperatures in Los Angeles, as the DLNM is more effective at controlling for variability at multiple temporal scales within the human mortality dataset.

Results: Significantly higher human mortality is uncovered in winter when dry tropical air is present in Los Angeles, particularly among those 65 years and older (p < 0.001). The DLNM reveals the relative risk of human mortality increases when above average temperatures are present. Results are especially pronounced for maximum and mean temperatures, along with total mortality and those 65 + .

Conclusions: The discovery of heat-related mortality in winter is a unique finding in the United States, and we recommend stakeholders consider warning and intervention techniques to mitigate the role of winter heat on human health in the County.

Keywords: Air masses; DLNM; Heat; Human mortality; Los Angeles; Winter heat waves.

Fig. 1

Average daily mortality for Los Angeles County from 1979 through 2010. A 15-day running mean was used to smooth the data

Fig. 2

Average daily mortality for Los Angeles County for the winters of 1988/1989 (a), 2009/2010 (b), 1980/1981 (c), 1987/1988 (d), 1992/1993 (e), 2007/2008 (f), and 1991/1992 (g). 15-day running means were used to smooth the data

Fig. 3

Total daily mortality in Los Angeles County (bars) and maximum daily temperature (line) from 27 December, 2002 through 14 January, 2003

Fig. 4

3-D plot of relative risk variation in mortality for those 65+ with daily maximum temperature at various lag times

Fig. 5

Relative risk for mortality of all ages at various daily (a) maximum, (b) mean, and (c) minimum temperatures. The red line is the predicted value and the gray shading represents the 95% confidence interval

Fig. 6

Relative risk for mortality of ages < 65 at various daily (a) maximum, (b) mean, and (c) minimum temperatures. The red line is the predicted value and the gray shading represents the 95% confidence interval

Fig. 7

Relative risk for mortality of ages 65+ at various daily (a) maximum, (b) mean, and (c) minimum temperatures. The red line is the predicted value and the gray shading represents the 95% confidence interval