Correspondence

Air Pollution and Life Expectancy

N Engl J Med 2009; 360:2032-2034May 7, 2009

Article

To the Editor:

In their study of the association between fine-particulate air pollution and life expectancy, Pope et al. (Jan. 22 issue)1 adjusted for changes in the prevalence of smoking in order to determine the contribution of reduced levels of air pollution to an improved life expectancy. As the authors observed previously,2 the effects of smoking on mortality dwarf those of air pollution, since being a current smoker increases the risk of lung cancer by 1480%, as compared with an increase of 8% per increase of 10 μg per cubic meter in fine-particulate pollution.

A potential arithmetic problem arises. Small errors in estimating either the extent of the change in the prevalence of smoking or the effect of smoking on life expectancy would greatly alter the calculation of the residual, relatively tiny effect of air pollution. Yet changes in tobacco use result in time-delayed effects on rates of cancer and death from cardiac causes that are difficult to estimate,3,4 with a 20-year lag between a decreased rate of smoking and a decreased rate of lung cancer.3 Thus, although the prevalence of smoking decreased, the proxy markers that were used for smoking (deaths from lung cancer or chronic obstructive pulmonary disease [COPD]) increased during the observation period, making them a curious choice. Similarly, would not health effects of air pollution be delayed by some incalculable time?

Michael R. Bubb, M.D.
University of Florida, Gainesville, FL 32610
bubbmr@medicine.ufl.edu

4 References

1. 1

   Pope CA III, Ezzati M, Dockery DW. Fine-particulate air pollution and life expectancy in the United States. N Engl J Med 2009;360:376-386
   Full Text | Web of Science | Medline

2. 2

   Pope CA III, Burnett RT, Thun MJ, et al. Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA 2002;287:1132-1141
   CrossRef | Web of Science | Medline

3. 3

   Thun MJ, Jemal A. How much of the decrease in cancer death rates in the United States is attributable to reductions in tobacco smoking? Tob Control 2006;15:345-347[Erratum, Tob Control 2006;15:488.]
   CrossRef | Web of Science | Medline

4. 4

   Barone-Adesi F, Vizzini L, Merletti F, Richiardi L. Short-term effects of Italian smoking regulation on rates of hospital admission for acute myocardial infarction. Eur Heart J 2006;27:2468-2472
   CrossRef | Web of Science | Medline

To the Editor:

Pope et al. attribute a 5-month gain in life expectancy at birth to improved ambient air quality, as measured during two periods, from 1979 through 1983 and from 1997 through 2001. This inconsistency in time scales for outcomes (over a lifetime) and exposures (over a 20-year period) exaggerates benefits; outcomes are overestimated, relevant changes in exposure are underestimated, or both. Long-term effects of air pollution on health tend to be manifested in middle age,1 during which the remaining life expectancy is approximately 10 to 20% of that at birth.2 The middle-aged subjects in the study were born between about 1900 and 1950; because of previous pollution abatement, their current pollution exposures were only a fraction of their earlier and lifetime exposures.3 Any effects on infant mortality must be evaluated against these earlier exposures.

In addition, this ecologic analysis considers only one pollutant and a surrogate measure of smoking status (concurrent lung cancer), thus neglecting differences in disease latency. It also ignores gradients in advances in medical care that contributed significantly to an improved adult life expectancy in the 1980s and 1990s.4 This inadequate treatment of confounders is a likely source of additional upward bias in the estimates of the effects of air pollution on longevity.5

Frederick W. Lipfert, Ph.D.
23 Carll Ct., Northport, NY 11768
flipfert@suffolk.lib.ny.us

5 References

1. 1

   Lipfert FW, Morris SC. Temporal and spatial relationships between age-specific mortality and ambient air quality in the United States: regression results for counties, 1960-97. Occup Environ Med 2002;59:156-174
   CrossRef | Web of Science | Medline

2. 2

   Anderson RN, DeTurk PB. United States life tables, 1999. Natl Vital Stat Rep 2002;50:33-33
   

3. 3

   Lipfert FW. Trends in airborne particulate matter in the United States. Appl Occup Environ Hyg 1998;13:370-384
   CrossRef

4. 4

   Hunink MG, Goldman L, Tosteson AN, et al. The recent decline in mortality from coronary heart disease, 1980-1990: the effect of secular trends in risk factors and treatment. JAMA 1997;277:535-542
   CrossRef | Web of Science | Medline

5. 5

   Bateson TF, Coull BA, Hubbell B, et al. Panel discussion review: session three -- issues involved in interpretation of epidemiologic analyses -- statistical modeling. J Expo Sci Environ Epidemiol 2007;17:Suppl 2:S90-S96
   CrossRef | Web of Science | Medline

Author/Editor Response

Bubb and Lipfert express their concerns regarding inadequate adjustment for smoking and other potential confounders in our study. Smoking has a substantial effect on life expectancy, but Bubb exaggerates the “potential arithmetic problem.” The excess risk of death from any cause would be approximately 100% for current smokers, as compared with an increase of approximately 6 to 15% for each increase of 10 μg per cubic meter in the concentration of particulate matter with an aerodynamic diameter less than or equal to 2.5 μm (PM_2.5).1 In populations ubiquitously exposed to moderately high levels of ambient air pollution (PM_2.5, 15 to 30 μg per cubic meter) but in which a fraction of people (15 to 20%) are smokers, the net population-level effects of smoking and pollution on life expectancy may be similar. We observed changes in life expectancy that were associated with changes in both air-pollution levels and proxy variables for smoking, which are independent of each other in multivariate regression models.

Lipfert suggests that inadequate treatment of confounders is a potential source of upward bias in the estimates of the effects of air pollution. Uncontrolled factors can result in upward or downward bias, depending on whether they are positively or negatively associated with air pollution. Direct measures of county-level access to advances in medical care are not available, but confounding by these variables is at least partially controlled by adjustment for socioeconomic and demographic variables. Associations between changes in life expectancy and air-pollution levels are statistically robust in analyses adjusted for changes in socioeconomic and demographic variables, providing little evidence of significant confounding. Nevertheless, we agree that potential residual confounding is a concern in this and other observational studies.

With regard to time scales of exposure, daily time-series, intervention, and prospective cohort studies provide substantial evidence that the associations between air-pollution levels and mortality have time scales of days to years. The time scale of exposure that captures most of the excess risk of death from any cause is recent exposure (over a period of up to a few years).1,2 Disease latency (and thus the relevant time scale) is relatively longer for lung cancer specifically. However, we also controlled for smoking, using rates of death from COPD and survey-based smoking prevalence at the metropolitan level. The estimated increases in life expectancy associated with declines in air-pollution levels were not highly sensitive to adjustment for any of the smoking-related variables. The data set used for the interactive graphic was redacted to make the presentation manageable. A full data set is available from Dr. Pope on request.

C. Arden Pope, III, Ph.D.
Brigham Young University, Provo, UT 84602-2363
cap3@byu.edu

Majid Ezzati, Ph.D.
Douglas W. Dockery, Sc.D.
Harvard School of Public Health, Boston, MA 02115

2 References

1. 1

   Pope CA III, Dockery DW. Health effects of fine particulate air pollution: lines that connect. J Air Waste Manag Assoc 2006;56:709-742
   Medline

2. 2

   Schwartz J, Coull B, Laden F, Ryan L. The effect of dose and timing of dose on the association between airborne particles and survival. Environ Health Perspect 2008;116:64-69
   CrossRef | Web of Science | Medline

Citing Articles (1)

Citing Articles

1. 1

   Arne Marian Leitte, Uwe Schlink, Olf Herbarth, Alfred Wiedensohler, Xiao-Chuan Pan, Min Hu, Birgit Wehner, Susanne Breitner, Annette Peters, H.-Erich Wichmann, Ulrich Franck. (2011) Associations between size-segregated particle number concentrations and respiratory mortality in Beijing, China. International Journal of Environmental Health Research1-15
   CrossRef