Years of potential life lost (YPLL) is a measure of premature mortality and can be used to describe the impact on society of death from a particular illness or injury [Kamel et al., 2012]. YPLL is used to help quantify economic loss owing to premature death [Gardner and Sanborn, 1990; Pham et al., 2009] and to target prevention programs and health care planning [Perloff et al., 1984; Gardner and Sanborn, 1990]. Also, YPLL can assist in disease surveillance, and in evaluation of mortality temporal trends and program intervention effectiveness [CDC, 1986]. YPLL emphasizes deaths occurring at younger age and YPLL that occur before age 65 years measures years of potential work tenure lost, on the assumption that these are a worker’s productive years [Wise et al., 1988; O’Lorcain et al., 2007].

Several diseases have been associated with exposure to asbestos fibers, including asbestosis, malignant mesothelioma, pleural plaques, and lung and other cancers [WHO, 1999; O’Reilly et al., 2007; Rudd, 2010; IARC, 2012]. Workers employed in asbestos mining and milling and those involved with manufacture, use, repair, removal, road maintenance, ship and boat building, construction, or demolition of asbestos-containing materials and products are at potential risk of exposure to asbestos [Bang et al., 2008]. Although asbestos is no longer mined in the United States, asbestos continues to be imported (approximately 1,060 metric tons in 2012) [US DI, 2013]. In 2012, the chloralkali industry consumed 67% of the imported asbestos to manufacture semipermeable diaphragms in the electrolytic cells. Thirty percent of imported asbestos was used in roofing products and the remaining 3% was used in coatings and compounds, and plastics [Virta, 2013]. Asbestos can be found in brake pads, vinyl tile, friction products, insulation, roofing materials, and other products [EPA, 2012a]. Moreover, a substantial amount of asbestos remains in buildings and eventually will be removed, either during remediation or renovations or demolition [EPA, 2012a]. It has been estimated that approximately 1.3 million workers in construction and general industry potentially are being exposed to asbestos during maintenance activities or remediation of buildings containing asbestos [Weeks and Christiani, 2005].

The U.S. Occupational Safety and Health Administration, the Mine Safety and Health Administration, and the Environmental Protection Agency have taken regulatory actions to control occupational exposure to asbestos by establishing a permissible exposure limit and banning some asbestos-containing products from the market [Martonik et al., 2001; MSHA, 2010; EPA, 2012b]. However, asbestos use has not been banned completely in the United States and asbestos-containing materials remain in place in structural materials and machinery. Non-regulated mineral fibers are causing asbestos-associated diseases as well. For example, in the United States mesothelioma was associated with commercial use of vermiculite mined in Libby, Montana [Duning et al., 2012] and asbestosis was associated with exposure to erionite which was used to pave some roads in North Dakota [Carbone et al., 2012]. Similar exposure to erionite was reported from Turkey [CDC, 2011]. Finally, estimated 41,000 persons were exposed to asbestos and other toxins during the rescue and clean-up efforts after the World Trade Center collapsed (http://www.asbestos.com/world-trade-center/). Thus, asbestos-associated deaths may continue to occur for decades to come [Nicholson et al., 1982; Antao et al., 2009; CDC, 2011].

Limited information on YPLL attributable to diseases associated with asbestos exposure is available [CDC, 2008a]. In this study, we examined YPLL attributable to mortality from asbestosis and malignant mesothelioma associated with asbestos exposure using 1999–2010 mortality data from the U.S. National Center for Health Statistics (NCHS).

We used the 1999–2010 NCHS data because the International Classification of Disease 10th revision (ICD-10) [WHO, 1992], which assigns a separate code for malignant mesothelioma, was adopted for coding deaths in the United States beginning in 1999. The most recent year for which NCHS mortality data were available was 2010.

Using ICD-10 codes for causes of death listed in the NCHS multiple-cause-of-death data files, we identified two non-overlapping groups of decedents with diseases attributable to asbestos exposure as follows: (1) Asbestosis [J61]: pneumoconiosis due to asbestos and other mineral fibers as underlying cause of death and death with underlying cause coded as J65 (pneumoconiosis associate with tuberculosis) or J92.0 (pleural plaques with presence of asbestos) with asbestosis as a contributing cause of death and (2) Malignant Mesothelioma: mesothelioma [C45] as underlying cause of death. Because of the occupational etiology and long latency of diseases associated with asbestos exposure [Rudd, 2010], analysis was restricted to deaths of persons aged ≥25 years.

During 1999–2010, asbestosis and malignant mesothelioma as the underlying cause of death were coded for 35,929 decedents (median age at death: 75 years), accounting for 427,005 YPLL (median per decedent: 10.0). The majority of decedents were male (29,819; 83.0%) and white (34,130; 95.0%), accounting for 345,856 (81.0%; median: 9.9) and 400,414 (93.8%; median: 9.9) YPLL, respectively (Table I). The median YPLL attributable to asbestosis and malignant mesothelioma for females (10.7) was significantly greater than that for males (9.9; P <0.001). The median YPLL attributable to asbestosis and malignant mesothelioma for blacks (12.1) was significantly greater than that for whites (9.9; P <0.001).

Overall, annual YPLL attributable to asbestosis and malignant mesothelioma did not change significantly (from 34,357 in 1999 to 35,818 in 2010; Table II). By state (Table III), the YPLL attributable to asbestosis and malignant mesothelioma ranged from 390 for 27 decedents in the District of Columbia (DC) to 41,384 for 3,407 decedents in California, while median asbestos-attributable YPLL per death ranged from 8.1 in North Dakota to 14.0 in DC. The age-adjusted YPLL rate per million was greatest in Maine (331). Ten states (California, Florida, Pennsylvania, Texas, New York, Illinois, Ohio, New Jersey, Michigan, and Washington) accounted for over half (52.9%) of the total YPLL attributable to asbestosis and malignant mesothelioma. These same 10 states accounted for 51.0% of total YPLL₆₅ attributable to asbestosis and malignant mesothelioma.

The YPLL is an impact measure that can help quantify the burden of social and economic loss from premature mortality [Gardner and Sanborn, 1990]. The use of premature mortality measures has recently become more common for targeting cancer because it can be used to describe premature death and potential economic loss to the society caused by cancer deaths and also it can enhance the decisions of agencies for evaluating cancer prevention programs [O’Lorcain et al., 2007; Pham et al., 2009; Kamel et al., 2012]. However, information on YPLL attributable to diseases associated with asbestos exposure is limited in the United States. Adoption of the ICD-10 for coding death data with a specific code for malignant mesothelioma in 1999 and availability of data for 1999–2010 provided data to address this gap. This study is the first to report on asbestos-attributable YPLL using national mortality data.

We found that deaths attributable to asbestosis and malignant mesothelioma resulted in an annual average of 35,600 YPLL and over 4,600 YPLL₆₅. Nearly 96% of the total YPLL₆₅ attributable to asbestosis and malignant mesothelioma was due to malignant mesothelioma. No significant temporal trend in malignant mesothelioma attributable YPLL or YPLL₆₅ was observed over the years for which data were available.

Because the latency period from asbestos exposure to the development of malignant mesothelioma ranges from 20 to 50 years, most malignant mesothelioma occurs at the age of over 45 [CDC, 2008b; Maskell and Millar, 2009]. The majority (85–87%) of these cases were likely exposed to asbestos at the workplace [Rudd, 2010]. Industries with known high exposure to asbestos in the past included shipbuilding, construction, building trades, automotive brake, and the manufacture of asbestos textiles [Maskell and Millar, 2009]. Malignant mesothelioma under the age of 45 is uncommon and indicates that exposure may have occurred in younger age, likely in non-occupational settings such as exposure to naturally occurring asbestos [Pan et al., 2005].

Although most non-occupational malignant mesothelioma is likely due to asbestos exposure, exposure to other factors such as erionite, and asbestos fiber-containing vermiculite and talc have also been associated with non-occupational malignant mesothelioma [Carbone et al., 2012; CDC, 2012; Duning et al., 2012; IARC, 2012]. Erionite deposits have been documented in ten western U.S. states (California, Nevada, North Dakota, South Dakota, Arizona, Colorado, Idaho, New Mexico, Utah, and Wyoming) and exposure to erionite was associated with interstitial and pleural changes in workers with prolonged occupational exposure to road gravel that contains fibrous erionite [Ryan et al., 2011]. Vermiculite has been documented to increase risk for the development of malignant mesothelioma in workers that processed vermiculite mined near Libby, Montana [Duning et al., 2012]. Finkelstein [2012] recently reported that workers in the New York State talc industry were at increased risk of malignant mesothelioma attributable to their inhalation of asbestos-containing talc in the mining and milling.

Most asbestosis deaths occur among persons aged 65–84 years and overall asbestosis mortality is not expected to decrease rapidly in the next 10–15 years in the United States [Antao et al., 2009]. In our study, small numbers of deaths among persons aged 25–64 were associated with asbestosis (n = 343) and lung cancer with asbestosis or pleural plaques with presence of asbestos (n = 200), accounting for 4% and 2% of the total asbestos-attributable YPLL₆₅, respectively. A median of 5.0 YPLL₆₅ per decedent was attributed to asbestosis and a median of 4.0 YPLL₆₅ per decedent was attributed to lung cancer associated with asbestos exposure indicating that, on average, decedents aged 25–64 years with asbestosis and those with lung cancer died at age 60 and 61 years, respectively.

A small number of pleural plaques (n = 27) with presence of asbestosis was identified in this study. Pleural plaques are associated with exposure to asbestos [WHO, 1999]. Peretz et al. [2008] reported that “take-home” exposure to asbestos brought by family members was associated with pleural plaques among family members.

There were some differences in YPLL and YPLL₆₅ by gender and race. Females had a higher median YPLL attributable to malignant mesothelioma than males, and blacks had a higher median YPLL attributable to malignant mesothelioma than whites, indicating that females are dying younger than males and that blacks are dying younger than whites. Additional studies are needed to ascertain which factors (e.g., age at exposure to asbestos, access to medical care) may be associated with women’s and blacks’ younger age at death from malignant mesothelioma.

In this study, nearly 60% of YPLL attributable to disease associated with asbestos exposure were calculated for decedents reported from coastal states. These findings may be explained, in part, by the historical exposure to asbestos in the ship and boat building and repairing industry in coastal states [Enterline and Henderson, 1987; Bang et al., 2008]. However, no information on decedents’ occupation was available to confirm these previous results.

Our method of computing YPLL to use 65 years is consistent with a study on premature mortality [Semerl and Sesok, 2002], coal workers’ pneumoconiosis-related YPLL [CDC, 2009], and asbestosis-related YPLL [CDC, 2008a]. For computation of YPLL to life expectancy, we used a life expectancy tables for ages 0–100 that was published by the National Center for Health Statistics. Pham et al. [2009] used life table methods for YPLL due to cancer in Japan. The method for calculating YPLL varies among authors, depending on cut-off age or life tables of different populations. YPLL varies by different occupational diseases. Zhong and Dehong [1995] reported that mean YPLL for asbestosis and silicosis among decedents aged 15–75 years was 14.3 and 21.9, respectively.

The findings in this study are subject to some limitations. First, YPLL and YPLL₆₅ may be undercounted because some deaths from asbestosis and malignant mesothelioma might have been miscoded. To our knowledge, no studies addressed the magnitude of potential bias associated with coding causes of deaths according to ICD-10 coding scheme. Second, because no separate ICD-10 code is assigned to cancers associated with asbestos exposure [IARC, 2012], we were not able to calculate the total YPLL and YPLL₆₅ attributable to these conditions. Steenland et al. [2003] estimated annual number of lung cancer deaths associated with industrial carcinogen exposures to range 6,800–17,000, using 1997 mortality data in the United States. Although the authors have considered asbestos exposure in their calculations, no discrete estimates of lung cancer associated with asbestos were reported. Third, the disability adjusted life year (DALY) could not be estimated in this study because there is no information on years lost due to disability for diseases associated with asbestos exposure that would allow for a more accurate description of the burden of diseases attributable to asbestos exposure [WHO, 2008]. In the future, the DALY could be estimated when information on disability for these diseases is available. Fourth, the state that issues a death certificate may not be the same state in which the decedent’s asbestos exposure occurred. Finally, no information on industry or occupation was available for analysis; thus, identification of specific industry and occupation associated with YPLL was not possible. The National Institute for Occupational Safety and Health has previously reported the industry and occupation information for malignant mesothelioma and asbestosis deaths from the 1999 NCHS mortality data which were included in this study [CDC, 2008b].

Asbestos-attributable diseases are associated with a substantial loss of potentially productive years of life. Future economic analysis might address the economic impact of asbestos-attributable diseases. The continuing occurrence of death from diseases associated with asbestos exposure underscores the need for maintaining asbestos exposure prevention and intervention efforts and for an ongoing surveillance to monitor temporal trends in these diseases.