Cardiovascular and Cerebrovascular Mortality Associated With Acute Exposure to PM2.5 in Mexico City
Abstract
Background and Purpose—
Acute exposure to particulate matter with aerodynamic diameter <2.5 μm (PM2.5) is associated with acute cardiovascular and cerebrovascular mortality. The aim of this study was to evaluate these associations with specific causes of cardiovascular and cerebrovascular mortality in Mexico City.
Methods—
We obtained daily mortality records for Mexico City from 2004 to 2013 for cardiovascular and cerebrovascular causes in people ≥25 and ≥65 years old. Exposure to PM2.5 was assessed with daily estimates from a new hybrid spatiotemporal model using satellite measurements of aerosol optical depth PM2.5 and compared to ground level PM2.5 measurements with missing data estimated with generalized additive models PM2.5. We fitted Poisson regression models with distributed lags for all mortality outcomes.
Results—
An increase of 10 µg/m3 in aerosol optical depth PM2.5 was associated with increased cardiovascular (1.22%; 95% confidence interval, 0.17–2.28) and cerebrovascular mortality (3.43%; 95% confidence interval, 0.10–6.28) for lag days 0 to 1 (lag 0–1). Stronger effects were identified for hemorrhagic stroke and people ≥65 years. Associations were slightly smaller using generalized additive models PM2.5.
Conclusions—
These results support the evidence that acute exposure to PM2.5 is associated with increased risk of specific cardiovascular and cerebrovascular mortality causes.
Introduction
Acute exposure to airborne particles with aerodynamic diameter ≤2.5 μm (PM2.5) can trigger cardiovascular and cerebrovascular mortality.1 In developing countries, the region with the highest burden of stroke,2 such evidence is limited3 and possibly related to lack of ground level monitoring of PM2.5.4
Mexico City, once considered the most polluted in the world, has improved its air quality because of various programs prioritizing public health.5 However, it continues to be among the most polluted cities in Latin America6 and over the past decades it has recorded an increasing rate of cardiovascular and cerebrovascular diseases, which are among the 5 leading causes of death.7
Despite the large body of scientific evidence about adverse health effects of particulate matter in Mexico City,8 epidemiological research about cardiovascular and cerebrovascular mortality associated with PM2.5 exposure remains limited. We therefore evaluated acute PM2.5 exposure associated with specific causes of cardiovascular and cerebrovascular mortality.
Methods
The data that support the findings of this study are available from the corresponding author on reasonable request. The present article also adheres to the American Heart Association Journals’ implementation of the Transparency and Openness Promotion guidelines.
Mortality Data and Exposure Assessment
We used an ecological design with public mortality records for Mexico City from 2004 to 2013. Our research was ruled exempt from human subjects review by the ethics board of the National Institute of Public Health of Mexico. Detailed methods are provided in the online-only Data Supplement. Deaths in people ≥25 and ≥65 years old, classified according to the International Classification of Diseases, Tenth Revision codes, were aggregated to obtain daily counts for specific mortality causes. To improve the quality of ischemic heart disease mortality data, we applied redistribution of misclassified cause of death by using the proportions estimated by Naghavi et al9 (Table I in the online-only Data Supplement). Daily citywide exposure to PM2.5 was assessed with estimates from a new hybrid spatiotemporal model using satellite measurements of aerosol optical depth (AOD-PM2.5) developed by Just et al.10 We also calculated daily PM2.5 averages from 3 monitoring stations of the Mexico City atmospheric monitoring system. On days with missing data, PM2.5 concentration was estimated with generalized additive models (GAM-PM2.5), with Methods described in the online-only Data Supplement.
Statistical Analysis
Associations were estimated using generalized linear models with Poisson regressions and distributed lags. Our base models included dummy variables for season of year, day of the week, penalized splines to address long-term trends in mortality and a natural cubic spline of apparent temperature. We alternatively included linear terms of NO2 and SO2 to investigate potential confounding effects of copollutants. Deviations from linearity in the concentration-response functions between PM2.5 and cardiovascular and cerebrovascular mortality featured comparison between models using linear PM2.5 and log-transformed PM2.5.
Results
The mean±SD for all deaths because of cardiovascular and cerebrovascular causes per day was 39±8.2. Ischemic heart disease mortality accounted for the highest proportion of all daily cardiovascular deaths (55%). Daily average counts for deaths because of ischemic and hemorrhagic stroke were 1.1±1.1 and 2.7±1.6, respectively. Daily averages for AOD-PM2.5 and GAM-PM2.5 were 24.4±8.2 µg/m3 and 25.9±10.3 µg/m3, respectively (Table II in the online-only Data Supplement describes mortality and environmental characteristics for the study period).
Table shows results on same day exposure (lag 0), cumulative effects over 2 (lag 0–1) and 6 days (lag 0–6) for AOD-PM2.5 and GAM-PM2.5. Significant mortality increments were observed for all cardiovascular and cerebrovascular mortality causes using AOD-PM2.5 (lag 0–1). No deviations from linearity were observed in the concentration-response functions in the associations assessed (exposure range of 3–99.5 µg/m3). Slightly lower risks were observed using GAM-PM2.5 compared with AOD-PM2.5, but no significant differences in health effects parameters and SEs were identified (Table IV in the online-only Data Supplement).
| Overall | AOD-PM2.5 | GAM-PM2.5 | |||
|---|---|---|---|---|---|
| Base Model | Fully Adjusted | Base Model | Fully Adjusted | ||
| % Change (95% CI) | % Change (95% CI) | % Change (95% CI) | % Change (95% CI) | ||
| Cardiovascular ≥25 y old | Lag 0 | 1.32 (0.50 to 2.15) | 1.02 (0.04 to 2.02) | 1.04 (0.38 to 1.70) | 0.76 (−0.02 to 1.55) |
| Lag 0–1 | 1.50 (0.64 to 2.38) | 1.22 (0.1 to 2.28) | 1.22 (0.52 to 1.92) | 0.94 (0.11 to 1.79) | |
| Lag 0–6 | 1.24 (0.25 to 2.24) | 0.99 (−0.10 to 2.10) | 0.77 (−0.09 to 1.64) | 0.59 (−0.34 to 1.53) | |
| Cerebrovascular ≥25 y old | Lag 0 | 2.92 (0.36 to 5.55) | 3.16 (0.13 to 6.27) | 2.05 (0.04 to 4.11) | 2.21 (−0.18 to 4.66) |
| Lag 0–1 | 3.15 (0.37 to 6.00) | 3.43 (0.10 to 6.28) | 2.48 (0.27 to 4.75) | 2.68 (0.01 to 5.42) | |
| Lag 0–6 | 2.54 (−0.60 to 5.78) | 2.76 (−0.68 to 6.33) | 2.07 (−0.59 to 4.80) | 2.22 (−0.69 to 5.22) | |
| Ischemic heart disease | Lag 0 | 0.61 (−0.47 to 1.70) | 0.51 (−0.77 to 1.82) | 0.70 (−0.16 to 1.57) | 0.66 (−0.36 to 1.70) |
| Lag 0–1 | 1.10 (−0.07 to 2.29) | 1.00 (−0.43 to 2.44) | 1.03 (0.09 to 1.98) | 1.00 (−0.14 to 2.14) | |
| Lag 0–6 | 0.75 (−0.55 to 2.07) | 0.67 (−0.77 to 2.13) | 0.47 (−0.65 to 1.61) | 0.44 (−0.79 to 1.70) | |
| Ischemic heart disease (improved by potential misclassification in death cause) | Lag 0 | 0.89 (−0.19 to 1.97) | 0.86 (−0.41 to 2.14) | 0.77 (−0.09 to 1.63) | 0.86 (−0.16 to 1.89) |
| Lag 0–1 | 1.37 (0.21 to 2.54) | 1.32 (−0.07 to 2.72) | 1.17 (0.24 to 2.10) | 1.20 (0.09 to 2.33) | |
| Lag 0–6 | 0.78 (−0.49 to 2.06) | 0.81 (−0.60 to 2.23) | 0.41 (−0.69 to 1.51) | 0.53 (−0.69 to 1.75) | |
| Ischemic stroke | Lag 0 | −0.86 (−5.64 to 4.15) | −0.60 (−6.60 to 5.78) | −0.75 (−4.45 to 3.09) | 1.01 (−3.43 to 5.66) |
| Lag 0–1 | 0.90 (−4.19 to 6.25) | 1.12 (−5.45 to 8.14) | 0.33 (−3.70 to 4.53) | 2.28 (−2.59 to 7.39) | |
| Lag 0–6 | 5.68 (−0.40 to 12.13) | 5.82 (−1.27 to 13.41) | 4.87 (−0.27 to 10.28) | 6.49 (0.77 to 12.54) | |
| Hemorrhagic stroke | Lag 0 | 3.80 (0.74 to 6.94) | 4.01 (0.37 to 7.77) | 2.72 (0.33 to 5.17) | 2.84 (−0.01 to 5.77) |
| Lag 0–1 | 3.16 (−0.08 to 6.50) | 3.36 (−0.58 to 7.46) | 2.72 (0.16 to 5.35) | 2.86 (−0.25 to 6.07) | |
| Lag 0–6 | 0.25 (−3.38 to 4.02) | 0.33 (−3.62 to 4.45) | −0.15 (−3.19 to 2.98) | −0.08 (−3.38 to 3.32) | |
| Cardiovascular ≥65 y old | Lag 0 | 1.65 (0.65 to 2.66) | 0.91 (−0.27 to 2.10) | 1.20 (0.41 to 1.99) | 0.53 (−0.41 to 1.48) |
| Lag 0–1 | 1.86 (0.80 to 2.92) | 1.05 (−0.20 to 2.32) | 1.41 (0.57 to 2.25) | 0.67 (−0.33 to 1.67) | |
| Lag 0–6 | 2.92 (1.72 to 4.13) | 2.29 (0.98 to 3.63) | 1.84 (0.83 to 2.86) | 1.22 (0.11 to 2.34) | |
| Cerebrovascular ≥65 y old | Lag 0 | 2.59 (−0.52 to 5.81) | 2.66 (−1.02 to 6.47) | 1.82 (−0.60 to 4.31) | 1.91 (−0.98 to 4.89) |
| Lag 0–1 | 4.01 (0.54 to 7.60) | 4.24 (0.12 to 8.52) | 3.19 (0.44 to 6.01) | 3.37 (0.09 to 6.76) | |
| Lag 0–6 | 4.54 (0.56 to 8.68) | 4.70 (0.39 to 9.19) | 3.87 (0.49 to 7.37) | 3.97 (0.32 to 7.75) | |
Discussion
This is the first study showing that acute exposure to PM2.5 is associated with specific cardiovascular and cerebrovascular mortality causes in Mexico City, the most populated city in North America. We found results consistent with previous studies with daily increments in cardiovascular mortality of 1%–2% for every 10 μg/m3 increase in PM2.5 (lag 0–1).11 For cerebrovascular mortality, our findings seemed larger than the 1.4% summary (95% confidence interval, 0.9%–1.9%) reported for cities in Europe, Asia, and North America (lag 0–1).12 Also, our results for people ≥65 years old are consistent with the conclusions from a meta-analysis, pointing out strong evidence of higher mortality risks in older populations associated with acute exposure to particulate matter than in younger populations.13
Most studies have presented combined results for stroke types associated with acute exposure to PM2.5; with weaker evidence for associations with different types of stroke. We observed greater effects for hemorrhagic stroke than for ischemic stroke. Possible explanations are higher frequency of hemorrhagic stroke observed in Mexico City and distribution of cofactors making its inhabitants more liable to suffer hemorrhagic stroke.
Limitations in our investigation include citywide exposure assessment to PM2.5. We may have failed to capture the spatiotemporal PM2.5 variability within Mexico City possibly biasing point estimates toward the null (Berkson type error).14 Also, the association between exposure to PM2.5 and onset of acute cardiovascular events might be subject to substantial underestimation related to exposure misclassification. We used date of death instead of time of symptom onset to assign exposure to PM2.5.
Time series studies in air pollution epidemiology generally rely on correct classification of death causes from government records. Even though we performed proportional redistribution of potentially misclassified death causes to ischemic heart disease mortality to reduce measurement error, there are other mortality outcomes subject to correction that were not addressed in our research.
PM2.5 toxicity depends on different factors besides concentration levels. Further research assessing the spatial distribution and composition of PM2.5 within Mexico City is needed to further refine our findings.
Sources of Funding
I. Gutiérrez-Avila was supported by the Fulbright and CONACyT student grants. Dr Just was supported by National Institutes of Health grants R00ES023450 and P30ES023515.
Disclosures
None.
Footnotes
References
- 1.
Brook RD, Rajagopalan S, Pope CA, Brook JR, Bhatnagar A, Diez-Roux AV ; American Heart Association Council on Epidemiology and Prevention, Council on the Kidney in Cardiovascular Disease, and Council on Nutrition, Physical Activity and Metabolism. Particulate matter air pollution and cardiovascular disease: an update to the scientific statement from the American Heart Association.Circulation. 2010; 121:2331–2378. doi: 10.1161/CIR.0b013e3181dbece1.LinkGoogle Scholar - 2.
Feigin VL, Roth GA, Naghavi M, Parmar P, Krishnamurthi R, Chugh S ; Global Burden of Diseases, Injuries and Risk Factors Study 2013 and Stroke Experts Writing Group. Global burden of stroke and risk factors in 188 countries, during 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013.Lancet Neurol. 2016; 15:913–924. doi: 10.1016/S1474-4422(16)30073-4.CrossrefMedlineGoogle Scholar - 3.
Ljungman PL, Mittleman MA . Ambient air pollution and stroke.Stroke. 2014; 45:3734–3741. doi: 10.1161/STROKEAHA.114.003130.LinkGoogle Scholar - 4.
Brauer M, Freedman G, Frostad J, van Donkelaar A, Martin RV, Dentener F . Ambient air pollution exposure estimation for the Global Burden of Disease 2013.Environ Sci Technol. 2016; 50:79–88. doi: 10.1021/acs.est.5b03709.CrossrefMedlineGoogle Scholar - 5.
Parrish DD, Singh HB, Molina L, Madronich S . Air quality progress in North American megacities: a review.Atmos Environ. 2011; 45:7015–7025.CrossrefGoogle Scholar - 6. World Health Organization (WHO). Data Summary: WHO’s Ambient Air Pollution Database-Update.2016. http://www.who.int/phe/health_topics/outdoorair/databases/cities/en/. Accessed May 31, 2018.Google Scholar
- 7. Institute for Health Metrics and Evaluation (IHME). GBD Compare Data Visualization.http://vizhub.healthdata.org/gbd-compare. Accessed January 11, 2018.Google Scholar
- 8.
Calderón-Garcidueñas L, Kulesza RJ, Doty RL, D’Angiulli A, Torres-Jardón R . Megacities air pollution problems: Mexico City metropolitan area critical issues on the central nervous system pediatric impact.Environ Res. 2015; 137:157–169. doi: 10.1016/j.envres.2014.12.012.CrossrefMedlineGoogle Scholar - 9.
Naghavi M, Makela S, Foreman K, O’Brien J, Pourmalek F, Lozano R . Algorithms for enhancing public health utility of national causes-of-death data.Popul Health Metr. 2010; 8:9. doi: 10.1186/1478-7954-8-9.CrossrefMedlineGoogle Scholar - 10.
Just AC, Wright RO, Schwartz J, Coull BA, Baccarelli AA, Tellez-Rojo MM . Using high-resolution satellite aerosol optical depth to estimate daily PM2.5 geographical distribution in Mexico City.Environ Sci Technol. 2015; 49:8576–8584. doi: 10.1021/acs.est.5b00859.CrossrefMedlineGoogle Scholar - 11.
Atkinson RW, Kang S, Anderson HR, Mills IC, Walton HA . Epidemiological time series studies of PM2.5 and daily mortality and hospital admissions: a systematic review and meta-analysis.Thorax. 2014; 69:660–665. doi: 10.1136/thoraxjnl-2013-204492.CrossrefMedlineGoogle Scholar - 12.
Wang Y, Eliot MN, Wellenius GA . Short-term changes in ambient particulate matter and risk of stroke: a systematic review and meta-analysis.J Am Heart Assoc. 2014; 3:e000983.LinkGoogle Scholar - 13.
Bell ML, Zanobetti A, Dominici F . Evidence on vulnerability and susceptibility to health risks associated with short-term exposure to particulate matter: a systematic review and meta-analysis.Am J Epidemiol. 2013; 178:865–876. doi: 10.1093/aje/kwt090.CrossrefMedlineGoogle Scholar - 14.
Zeger SL, Thomas D, Dominici F, Samet JM, Schwartz J, Dockery D . Exposure measurement error in time-series studies of air pollution: concepts and consequences.Environ Health Perspect. 2000; 108:419–426.CrossrefMedlineGoogle Scholar
