Smoke Exposure and COVID-19 in California
Interactive Exploration of Association between Smoke Exposure and COVID-19 in 2020
Please navigate to the link in order to explore the incidence of COVID-19 cases and deaths for each month of 2020 as well as the smoke exposure for particulate matter (PM) of 2.5 µm in California in 2020.
You can select each month of 2020 for COVID-19 cases and deaths and it will display the incidence per 10,000 persons for each county for that month. Below the COVID-19 map is an interactive map of smoke exposure. This displays the PM 2.5 µm smoke exposure in each county of California in 2020. You can pick each month to display to correspond to the COVID-19 maps.
Since the influence and timing of smoke exposure leading to a respiratory infection is not clear, this lets you interact with different time points between smoke exposure and COVID-19. For example, if there is an incubation period between smoke exposure that lasts several weeks or months you can look at COVID-19 months after the smoke exposure. This might be expected if smoke exposure causes lung injury that pre-disposes individuals to an increased risk of COVID-19
Please Click on the links below to explore the assoications between Smoke Exposure and COVID-19
Smoke Exposure and COVID-19 in 2020
Smoke Exposure and COVID-19 in 2021
Smoke Exposure and COVID-19 in 2022
Explore the Associations of Population and Demographic Data
Population Map
According to the American Community Survey (ACS) data, California had an estimated population of over 39 million in 2020, making it the most populous state in the United States. The population is diverse, with individuals from various ethnic and racial backgrounds. About 60% of the population resides in urban areas, with the largest cities being Los Angeles, San Diego, and San Jose. The median household income is $80,440. The education level is diverse, with approximately 31% of the population holding a bachelor’s degree or higher. The following map shows the population of each county on a log base 10 scale. You can hover over each county to view additional demographic data.
Demographic Maps
Median Income
Rate of Workers Predominantly Outdoors
Since smoke is likely to affect workers who spend most of their time outdoors, I wanted to look at the rate of outdoor laborers in California. The variables used to determine outdoor laborers compared to all laborers were “B24011_031” and “B24011_034” from the ACS data. These two variables represent the number of workers 16 years and over who worked in farming, fishing, and forestry occupations, specifically those who worked on a farm, ranch, or in an orchard (B24011_031) or those who worked in other farming, fishing, and forestry occupations (B24011_034). These two variables were selected and summed to calculate the total number of outdoor laborers per county. The total number of laborers per county was also calculated by summing all the variables starting with “B24011”. The outdoor laborer rate was then determined by dividing the total number of outdoor laborers by the total number of laborers in each county, and multiplying by 100 to express it as a percentage.
Mask Use Survey
This data was taken from the New York Times github COVID-19 page. I am displaying the percent of persons who rated “ALWAYS” wearing a mask, since this was the most common survey response. The other responses can be seen in the label when you hover over the county.
Specifically from the READme.md description at https://github.com/nytimes/covid-19-data/tree/master/mask-use:
“This data comes from a large number of interviews conducted online by the global data and survey firm Dynata at the request of The New York Times. The firm asked a question about mask use to obtain 250,000 survey responses between July 2 and July 14, enough data to provide estimates more detailed than the state level. (Several states have imposed new mask requirements since the completion of these interviews.)
Specifically, each participant was asked: How often do you wear a mask in public when you expect to be within six feet of another person?
This survey was conducted a single time, and at this point we have no plans to update the data or conduct the survey again.”
Environmental Maps
Temperature
Precipitation
Elevation
Statistical Analysis of Smoke Association with COVID-19 Cases Lagged 1, 2, and 3 Months
2020
Univariate Linear Mixed Effects Modelling
Cases
Treating Smoke as Continuous
| estimate | lower_95 | upper_95 | p_value |
|---|---|---|---|
| 20.88 | 16.29 | 25.47 | 0.0000 |
| 9.67 | 4.53 | 14.81 | 0.0002 |
| -8.80 | -14.86 | -2.74 | 0.0046 |
Deaths
| estimate | lower_95 | upper_95 | p_value |
|---|---|---|---|
| 0.25 | 0.15 | 0.35 | 0.0000 |
| 0.01 | -0.10 | 0.12 | 0.9172 |
| -0.15 | -0.27 | -0.03 | 0.0099 |
Multivariate Linear Mixed Effects Modelling
Linear mixed model fit by REML. t-tests use Satterthwaite's method [
lmerModLmerTest]
Formula: covid_incidence_1mo ~ smoke + median_income + outdoor_laborer_rate +
avg_temp + precip + ALWAYS + (1 | NAME)
Data: data
REML criterion at convergence: 7176.7
Scaled residuals:
Min 1Q Median 3Q Max
-2.4982 -0.5898 -0.1321 0.3305 6.2005
Random effects:
Groups Name Variance Std.Dev.
NAME (Intercept) 75265 274.3
Residual 636903 798.1
Number of obs: 445, groups: NAME, 58
Fixed effects:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) 510.093 61.914 93.682 8.239 1.03e-12 ***
smoke 21.556 2.354 410.845 9.157 < 2e-16 ***
median_income -89.051 70.502 60.091 -1.263 0.211429
outdoor_laborer_rate -39.157 68.590 66.620 -0.571 0.570005
avg_temp 178.852 67.501 59.569 2.650 0.010303 *
precip -237.466 66.659 61.445 -3.562 0.000717 ***
ALWAYS -39.406 62.418 56.789 -0.631 0.530356
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Correlation of Fixed Effects:
(Intr) smoke mdn_nc otdr__ avg_tm precip
smoke -0.529
median_incm -0.026 0.025
otdr_lbrr_r 0.040 -0.112 0.579
avg_temp -0.060 0.110 0.145 0.062
precip -0.014 0.028 0.009 -0.134 0.480
ALWAYS 0.002 -0.001 -0.286 0.030 -0.242 0.169Linear mixed model fit by REML. t-tests use Satterthwaite's method [
lmerModLmerTest]
Formula: death_incidence_1mo ~ smoke + median_income + outdoor_laborer_rate +
avg_temp + precip + ALWAYS + (1 | NAME)
Data: data
REML criterion at convergence: 3780.1
Scaled residuals:
Min 1Q Median 3Q Max
-2.9921 -0.4638 -0.1661 0.2418 8.2021
Random effects:
Groups Name Variance Std.Dev.
NAME (Intercept) 34.23 5.851
Residual 272.03 16.493
Number of obs: 445, groups: NAME, 58
Fixed effects:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) 7.47503 1.29405 94.10307 5.776 9.80e-08 ***
smoke 0.27043 0.04867 410.80054 5.556 4.97e-08 ***
median_income -3.33210 1.47778 61.03541 -2.255 0.0277 *
outdoor_laborer_rate -0.86102 1.43669 67.45460 -0.599 0.5510
avg_temp 5.93542 1.41497 60.46422 4.195 9.08e-05 ***
precip -1.92430 1.39702 62.33153 -1.377 0.1733
ALWAYS 2.45165 1.30886 57.72457 1.873 0.0661 .
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Correlation of Fixed Effects:
(Intr) smoke mdn_nc otdr__ avg_tm precip
smoke -0.523
median_incm -0.026 0.024
otdr_lbrr_r 0.038 -0.111 0.579
avg_temp -0.058 0.108 0.145 0.062
precip -0.014 0.027 0.009 -0.135 0.479
ALWAYS 0.002 -0.001 -0.286 0.030 -0.242 0.170Linear mixed model fit by REML. t-tests use Satterthwaite's method [
lmerModLmerTest]
Formula: covid_incidence_1mo ~ smoke + median_income + outdoor_laborer_rate +
avg_temp + precip + ALWAYS + (1 | NAME)
Data: data
REML criterion at convergence: 7176.7
Scaled residuals:
Min 1Q Median 3Q Max
-2.4982 -0.5898 -0.1321 0.3305 6.2005
Random effects:
Groups Name Variance Std.Dev.
NAME (Intercept) 75265 274.3
Residual 636903 798.1
Number of obs: 445, groups: NAME, 58
Fixed effects:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) 510.093 61.914 93.682 8.239 1.03e-12 ***
smoke 21.556 2.354 410.845 9.157 < 2e-16 ***
median_income -89.051 70.502 60.091 -1.263 0.211429
outdoor_laborer_rate -39.157 68.590 66.620 -0.571 0.570005
avg_temp 178.852 67.501 59.569 2.650 0.010303 *
precip -237.466 66.659 61.445 -3.562 0.000717 ***
ALWAYS -39.406 62.418 56.789 -0.631 0.530356
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Correlation of Fixed Effects:
(Intr) smoke mdn_nc otdr__ avg_tm precip
smoke -0.529
median_incm -0.026 0.025
otdr_lbrr_r 0.040 -0.112 0.579
avg_temp -0.060 0.110 0.145 0.062
precip -0.014 0.028 0.009 -0.134 0.480
ALWAYS 0.002 -0.001 -0.286 0.030 -0.242 0.169Table: Model 1: Covid Incidence|Variable | Estimate| Std. Error| p-value| 95% CI Lower| 95% CI Upper||:--------------------|--------:|----------:|-------:|------------:|------------:||(Intercept) | 510.09| 61.91| 0.00| 392.89| 628.80||smoke | 21.56| 2.35| 0.00| 16.82| 26.07||median_income | -89.05| 70.50| 0.21| -223.05| 44.17||outdoor_laborer_rate | -39.16| 68.59| 0.57| -168.83| 91.43||avg_temp | 178.85| 67.50| 0.01| 50.64| 306.43||precip | -237.47| 66.66| 0.00| -364.30| -111.54||ALWAYS | -39.41| 62.42| 0.53| -157.07| 79.85||sd__(Intercept) | 274.35| NA| NA| NA| NA||sd__Observation | 798.06| NA| NA| NA| NA|
Table: Model 2: Death Incidence|Variable | Estimate| Std. Error| p-value| 95% CI Lower| 95% CI Upper||:--------------------|--------:|----------:|-------:|------------:|------------:||(Intercept) | 7.48| 1.29| 0.00| 5.02| 9.95||smoke | 0.27| 0.05| 0.00| 0.17| 0.36||median_income | -3.33| 1.48| 0.03| -6.13| -0.53||outdoor_laborer_rate | -0.86| 1.44| 0.55| -3.58| 1.87||avg_temp | 5.94| 1.41| 0.00| 3.25| 8.62||precip | -1.92| 1.40| 0.17| -4.57| 0.73||ALWAYS | 2.45| 1.31| 0.07| -0.03| 4.93||sd__(Intercept) | 5.85| NA| NA| NA| NA||sd__Observation | 16.49| NA| NA| NA| NA|
Table: Model 1: Covid Incidence Controlled for Month|Variable | Estimate| Std. Error| p-value| 95% CI Lower| 95% CI Upper||:--------------------|--------:|----------:|-------:|------------:|------------:||(Intercept) | -86.95| 116.36| 0.46| -308.06| 139.85||smoke | -3.37| 3.30| 0.31| -9.75| 3.00||median_income | -120.21| 69.40| 0.09| -251.31| 10.73||outdoor_laborer_rate | -12.64| 66.82| 0.85| -138.75| 113.62||avg_temp | 162.89| 66.43| 0.02| 37.28| 288.18||precip | -264.31| 65.25| 0.00| -387.64| -141.22||ALWAYS | -47.34| 61.47| 0.44| -163.26| 69.00||MonthAugust | 1414.10| 160.30| 0.00| 1100.49| 1721.20||MonthDecember | 1442.20| 145.03| 0.00| 1157.70| 1719.99||MonthFebruary | 162.83| 670.17| 0.81| -1146.79| 1447.86||MonthJuly | 474.01| 140.10| 0.00| 200.25| 743.13||MonthJune | 199.76| 146.56| 0.17| -85.72| 482.02||MonthMarch | 129.23| 257.98| 0.62| -375.28| 623.99||MonthMay | 71.12| 156.46| 0.65| -232.98| 372.88||MonthNovember | 968.06| 137.69| 0.00| 697.70| 1231.61||MonthOctober | 1120.74| 146.70| 0.00| 833.19| 1401.64||MonthSeptember | 1897.51| 184.64| 0.00| 1537.13| 2251.56||sd__(Intercept) | 320.75| NA| NA| NA| NA||sd__Observation | 637.09| NA| NA| NA| NA|2020, Assessing Smoke Data Without Removing Non-Smoke Factors
county_code NAME year month
Min. : 1.00 Length:612 Min. :2020 Min. : 1.00
1st Qu.: 29.00 Class :character 1st Qu.:2020 1st Qu.: 3.75
Median : 59.00 Mode :character Median :2020 Median : 6.50
Mean : 57.98 Mean :2020 Mean : 6.50
3rd Qu.: 85.00 3rd Qu.:2020 3rd Qu.: 9.25
Max. :113.00 Max. :2020 Max. :12.00
monthly_avg_smoke daily_aqi_value date
Min. : 1.083 Min. : 0.0 Length:612
1st Qu.: 5.331 1st Qu.: 19.0 Class :character
Median : 8.165 Median : 32.5 Mode :character
Mean : 12.125 Mean : 42.8
3rd Qu.: 14.241 3rd Qu.: 55.0
Max. :165.134 Max. :187.0 Linear mixed model fit by REML. t-tests use Satterthwaite's method [
lmerModLmerTest]
Formula: covid_incidence_1mo ~ monthly_avg_smoke + covid_incidence_2mo +
covid_incidence_3mo + (1 | NAME)
Data: data
REML criterion at convergence: 5791.9
Scaled residuals:
Min 1Q Median 3Q Max
-4.3227 -0.4895 -0.2230 0.2980 6.7034
Random effects:
Groups Name Variance Std.Dev.
NAME (Intercept) 13866 117.8
Residual 453615 673.5
Number of obs: 365, groups: NAME, 51
Fixed effects:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) 353.55636 60.89715 71.83484 5.806 1.61e-07 ***
monthly_avg_smoke 9.03600 2.57825 360.20076 3.505 0.000515 ***
covid_incidence_2mo 0.79960 0.05618 351.15564 14.234 < 2e-16 ***
covid_incidence_3mo -0.26927 0.05310 272.56788 -5.071 7.33e-07 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Correlation of Fixed Effects:
(Intr) mnth__ cvd__2
mnthly_vg_s -0.570
cvd_ncdnc_2 -0.109 -0.314
cvd_ncdnc_3 -0.210 0.215 -0.712| estimate | lower_95 | upper_95 | p_value |
|---|---|---|---|
| 27.47 | 22.19 | 32.75 | 0.0000 |
| 18.86 | 13.16 | 24.56 | 0.0000 |
| 0.04 | -6.59 | 6.67 | 0.9903 |
2021
Univariate Linear Mixed Effects Modelling
Cases
Treating Smoke Exposre as Continuous
| estimate | lower_95 | upper_95 | p_value |
|---|---|---|---|
| -0.12 | -7.32 | 7.08 | 0.9730 |
| 0.01 | -7.44 | 7.46 | 0.9978 |
| -0.01 | -7.73 | 7.71 | 0.9971 |
Deaths
| estimate | p_value |
|---|---|
| -0.0034 | 0.9579 |
| -0.0012 | 0.9862 |
| -0.0030 | 0.9650 |
Multivarirate Linear Mixed Regression Modelling
Table: Model 1: Covid Incidence
|Variable | Estimate| Std. Error| p-value| 95% CI Lower| 95% CI Upper|
|:--------------------|--------:|----------:|-------:|------------:|------------:|
|(Intercept) | -5.98| 1434.12| 1.00| -2808.15| 2796.19|
|monthly_avg_smoke | -0.19| 3.81| 0.96| -7.64| 7.26|
|median_income | 0.00| 0.00| 1.00| -0.01| 0.01|
|outdoor_laborer_rate | 0.24| 16.92| 0.99| -32.82| 33.29|
|avg_temp | -0.05| 16.66| 1.00| -32.61| 32.51|
|precip | -0.09| 5.27| 0.99| -10.38| 10.20|
|Month | 1.45| 16.16| 0.93| -30.14| 33.03|
|sd__(Intercept) | 0.00| NA| NA| NA| NA|
|sd__Observation | 1469.18| NA| NA| NA| NA|
Table: Model 2: Death Incidence
|Variable | Estimate| Std. Error| p-value| 95% CI Lower| 95% CI Upper|
|:--------------------|--------:|----------:|-------:|------------:|------------:|
|(Intercept) | -0.41| 25.38| 0.99| -50.00| 49.18|
|monthly_avg_smoke | -0.01| 0.07| 0.94| -0.14| 0.13|
|median_income | 0.00| 0.00| 1.00| 0.00| 0.00|
|outdoor_laborer_rate | 0.01| 0.30| 0.98| -0.58| 0.59|
|avg_temp | 0.00| 0.29| 1.00| -0.57| 0.58|
|precip | 0.00| 0.09| 0.99| -0.18| 0.18|
|Month | 0.04| 0.29| 0.90| -0.52| 0.59|
|sd__(Intercept) | 0.00| NA| NA| NA| NA|
|sd__Observation | 26.00| NA| NA| NA| NA|2022
Univariate Linear Mixed Effects Modelling
Cases
Treating Smoke Exposure as Continuous
| estimate | p_value |
|---|---|
| 11.5873 | 0.7071 |
| 17.4766 | 0.5938 |
| 15.6723 | 0.6459 |
Deaths
| estimate | p_value |
|---|---|
| 0.1001 | 0.7953 |
| 0.1251 | 0.7600 |
| 0.1348 | 0.7519 |
Multivariate Linear Mixed Regression Modelling
Table: Model 1: Covid Incidence
|Variable | Estimate| Std. Error| p-value| 95% CI Lower| 95% CI Upper|
|:--------------------|--------:|----------:|-------:|------------:|------------:|
|(Intercept) | 2268.70| 4519.52| 0.62| -6438.31| 11024.32|
|monthly_avg_smoke | 0.47| 31.02| 0.99| -61.81| 60.30|
|median_income | -0.01| 0.01| 0.30| -0.03| 0.01|
|outdoor_laborer_rate | 37.57| 58.43| 0.52| -75.02| 150.25|
|avg_temp | 6.45| 52.13| 0.90| -94.54| 106.89|
|precip | -31.73| 23.13| 0.17| -76.43| 12.85|
|sd__(Intercept) | 862.62| NA| NA| NA| NA|
|sd__Observation | 3483.62| NA| NA| NA| NA|
Table: Model 2: Death Incidence
|Variable | Estimate| Std. Error| p-value| 95% CI Lower| 95% CI Upper|
|:--------------------|--------:|----------:|-------:|------------:|------------:|
|(Intercept) | -32.95| 57.83| 0.57| -144.29| 78.65|
|monthly_avg_smoke | -0.06| 0.39| 0.88| -0.83| 0.68|
|median_income | 0.00| 0.00| 0.12| 0.00| 0.00|
|outdoor_laborer_rate | 0.85| 0.75| 0.26| -0.60| 2.30|
|avg_temp | 0.84| 0.67| 0.21| -0.45| 2.12|
|precip | -0.09| 0.30| 0.76| -0.66| 0.48|
|sd__(Intercept) | 11.47| NA| NA| NA| NA|
|sd__Observation | 43.15| NA| NA| NA| NA|