Report on the Environment
Cardiovascular Disease Prevalence and Mortality
What are the trends in human disease and conditions for which environmental contaminants may be a risk factors including across population subgroups and geographic regions?
The above question pertains to all 'Human Disease and Conditions ' Indicators, however, the information on these pages (overview, graphics, references and metadata) relates specifically to "Cardiovascular Disease Prevalence and Mortality". Use the right side drop list to view the other related indicators on this question.
- Asthma Prevalence
- Birth Defects Prevalence and Mortality
- Cancer Incidence
- Cardiovascular Disease Prevalence and Mortality
- Childhood Cancer Incidence
- Chronic Obstructive Pulmonary Disease Prevalence and Mortality
- Infectious Diseases Associated with Environmental Exposures or Conditions
- Low Birthweight
- Preterm Delivery
Click to enlarge exhibit
The broad category of cardiovascular disease (CVD) includes any disease involving the heart and blood vessels. Coronary heart disease, cerebrovascular disease (commonly known as stroke), and hypertension are the major cardiovascular diseases (American Heart Association, 2007). In addition to being a major risk factor for heart disease and stroke, hypertension is a commonly diagnosed disease that can also lead to kidney damage and other health problems. Obesity, physical inactivity, and sodium intake are all important risk factors for hypertension (NIH, 2004). Since 1900, CVD has been the leading cause of death in the U.S. every year except 1918 (American Heart Association, 2007) (General Mortality indicator). The U.S. age-adjusted mortality rate for CVD reached a peak in 1950 (CDC, 1999). Between 1950 and 1999, the age-adjusted mortality rate for CVD declined 60 percent. The major risk factors for CVD include tobacco use, high blood pressure, high blood cholesterol, diabetes, physical inactivity, and poor nutrition (CDC, 2004; American Heart Association, 2007).
Environmental exposures may also play a role in CVD morbidity and mortality independent of other risk factors. However, susceptible populations such as the elderly and other high-risk populations may be most impacted. For example, studies have shown exposure to ambient airborne particulate matter to be associated with increased hospitalizations and mortality among older individuals, largely due to cardiopulmonary and cardiovascular disease (U.S. EPA, 2004). Environmental tobacco smoke (ETS) may also contribute to CVD. Although the smoke to which a nonsmoker is exposed is less concentrated than that inhaled by smokers, research has demonstrated increased cardiovascular-related health risks associated with ETS (State of California, 2005).
This indicator presents U.S. adult (age 18 and older) prevalence rates for heart disease (all types), coronary heart disease, stroke, and hypertension; and mortality rates for CVD as a whole as well as coronary heart disease (including myocardial infarction), stroke, and hypertension. CVD prevalence data were compiled between 1997 and 2009 from the National Health Interview Survey (NHIS), conducted by the Centers for Disease Control and Prevention’s (CDC’s) National Center for Health Statistics (NCHS). The NHIS is the principal source of information on the health of the civilian non-institutionalized population of the U.S. and since 1960 has been one of the major data collection programs of NCHS. CVD prevalence is based on the number of adults who reported that they had ever been told by a doctor or other health practitioner that they had a specified CVD. Mortality data (all ages) were compiled between 1979 and 2007 using the National Vital Statistics System (NVSS), maintained by NCHS. The NVSS registers virtually all deaths and births nationwide, with data coverage from 1933 to 2007 and from all 50 states and the District of Columbia.
Among adults 18 years and older, the prevalence of heart disease and stroke between 1997 and 2009 has remained essentially the same (Exhibit 5-23). In contrast, the prevalence of hypertension has shown a general increase from 191 cases per 1,000 in 1997 to 249 cases per 1,000 in 2009.
Gender, race, and age differences in CVD prevalence exist. The prevalence of coronary heart disease is consistently higher among males than among females (81.1 cases per 1,000 for men compared with 49.6 cases per 1,000 for women in 2009). In contrast, hypertension is more prevalent among women (252.3 cases per 1,000 for women compared with 245.2 per 1,000 for men in 2008); however, the gap is narrowing as rates for men have been increasing at a faster pace over time compared to women. Among the racial groups reported, American Indians and Alaska Natives had the highest prevalence of coronary heart disease between 1999 and 2001. Between 2002 and 2009, however, the rates of coronary heart disease in this population exhibited much greater fluctuations ranging from 23.5 per 1,000 (2002) to 69.3 per 1,000 (2003). In 2009, whites had the highest prevalence of coronary heart disease (67.9 cases per 1,000), followed by blacks or African Americans (59.8 cases per 1,000), Asians (32.2 cases per 1,000), and American Indians and Alaska Natives (31.8 cases per 1,000). Between 1999 and 2009, Asians consistently had the lowest prevalence of stroke (9.8 cases per 1,000 in 2009) and hypertension (175.8 cases per 1,000 in 2009) among the racial groups reported. In addition, the Hispanic or Latino population had a consistently lower prevalence of the major CVD-related diseases compared with the non-Hispanic or Latino population from 1999-2009, the period for which these data are available. For example, in 2009, prevalence in Hispanics or Latinos was lower than in non-Hispanics or Latinos for coronary heart disease (37.9 versus 69.1 cases per 1,000, respectively), hypertension (162.5 versus 262.6 cases per 1,000, respectively), and stroke (13.3 versus 28.5 cases per 1,000, respectively). (Data not shown.)
In 1998, the national age-adjusted CVD mortality rate (all types) was 352.0 per 100,000 compared to a rate of 541.0 per 100,000 in 1980 (Exhibit 5-24). This decline continues after 1999, with the rate dropping from 349.3 per 100,000 in 1999 to 249.9 per 100,000 in 2007. Mortality rates for coronary heart disease, stroke, and myocardial infarction—subcategories of CVD—have also declined between 1979 and 1998. The age-adjusted coronary heart disease mortality rate ranged from 345.2 per 100,000 in 1980 to 197.1 per 100,000 in 1998. For stroke mortality, the age-adjusted rate ranged from 97.1 per 100,000 in 1979 to 59.3 per 100,000 in 1998. The age-adjusted mortality rates for myocardial infarction ranged from 157.9 in 1979 to 76.0 per 100,000 in 1998. The decline in mortality from these three CVD subgroups continued to be observed between 1999 and 2007. The age-adjusted mortality rates for coronary heart disease, stroke, and myocardial infarction in 2007 were 126.0, 42.2, and 41.4 per 100,000, respectively, compared to 194.6, 61.6, and 73.2 per 100,000, respectively, in 1999. In contrast, mortality attributed to hypertension has slightly increased between 1999 and 2007 from 15.8 per 100,000 to 17.8 per 100,000.
Both coronary heart disease and stroke mortality have been declining over time in each of the 10 EPA Regions (Exhibits 5-25 and 5-26). In 1979, coronary heart disease and stroke age-adjusted mortality rates ranged from 285.6 (Region 10) to 401.9 (Region 2) per 100,000 and 80.3 (Region 2) to 111.4 (Region 4) per 100,000, respectively. In 1998, coronary heart disease and stroke mortality rates ranged from 145.6 (Region 8) to 233.2 (Region 2) per 100,000 and 43.2(Region 2) to 68.5 (Region 10) per 100,000, respectively. The decreases in coronary heart disease and stroke mortality also appear to continue in the 1999-2007 period across all EPA regions. In 1999, coronary heart disease and stroke age-adjusted mortality rates ranged from 140.4 (Region 8) to 234.8 (Region 2) per 100,000 and 43.8 (Region 2) to 72.8 (Region 10) per 100,000, respectively. In 2007, coronary heart disease and stroke mortality rates ranged from 95.2 (Region 8) to 159.2 (Region 2) per 100,000 and 30.3 (Region 2) to 49.8 (Region 6) per 100,000, respectively.
Differences exist in CVD mortality rates among gender, racial, and age groups. For example, in 2007, those age 65 and older had the highest CVD (all types), coronary heart disease, and stroke mortality (1,633, 818.6, and 287.6 per 100,000, respectively). For the same year, the age-adjusted CVD, coronary heart disease, and stroke mortality rates for those 45 to 64 years of age were 161.1, 89.2, and 21.4 per 100,000, respectively. Notable differences in CVD (all types) and, specifically, coronary heart disease mortality rates exist between males and females, but not for stroke mortality. Coronary heart disease mortality among males in 2007 was 165.4 per 100,000, compared to 95.7 per 100,000 for females. In 2007, black or African American males had the highest CVD mortality rate (403.5 per 100,000) compared to white males (292.7 per 100,000), black or African American females (284.2 per 100,000), and white females (204.6 per 100,000). The lowest CVD mortality rates were observed for Asian or Pacific Islander females (125.5 per 100,000) followed by Asian or Pacific Islander males (174.8 per 100,000). (Data not shown.)
- Prevalence data reported in the NHIS are based on self-reported responses to specific questions pertaining to CVD-related illnesses, and are subject to the biases associated with self-reported data. Self-reported data can underestimate the disease prevalence being measured if, for whatever reason, the respondent is not fully aware of his/her condition.
- All prevalence data are based on crude rates and are not age-adjusted, as CDC did not report age-adjusted data prior to 2002 in the data sources used for this indicator. Because the possible influence of subgroup age structure has not been removed, the reported disease prevalence rates across time or within different race and gender subgroups may reflect differences in the age distribution of the populations being compared.
- For one or more years for which data are presented, coronary heart disease and stroke prevalence rates presented for Native Americans and Alaska Natives have a relative standard error of greater than 30 percent. In addition, stroke prevalence rates for one or more years for which data are presented for Asians have a relative standard error of greater than 30 percent. As such, these rates should be used with caution as they do not meet the standard of reliability or precision.
- CVD mortality rates are based on underlying cause of death as entered on a death certificate by a physician. Some individuals may have had competing causes of death. “When more than one cause or condition is entered by the physician, the underlying cause is determined by the sequence of conditions on the certificate, provisions of the ICD [International Classification of Diseases], and associated selection rules and modifications” (CDC, n.d.). Consequently, some misclassification of reported mortality might occur in individuals with competing causes of death, as well as the possible underreporting of CVD as the cause of death.
- The International Classification of Diseases 9th Revision (ICD-9) codes were used to specify underlying cause of death for years 1979-1998. Beginning in 1999, cause of death is specified with the International Classification of Diseases 10th Revision (ICD-10) codes. The two revisions differ substantially, and to prevent confusion about the significance of any specific disease code, data queries are separate.
CVD prevalence data were obtained from annual reports published by NCHS (NCHS, 1999-2005, 2006a,b, 2007, 2008, 2009a,b, 2010), which summarize health statistics compiled from the NHIS (http://www.cdc.gov/nchs/products/pubs/pubd/series/ser.htm). CVD mortality statistics were obtained from CDC’s “compressed mortality” database, accessed through CDC WONDER (CDC, 2011) (http://wonder.cdc.gov/mortSQL.html). EPA Regional mortality statistics were generated by combining and age-adjusting state-by-state totals for each EPA Region using data from CDC WONDER.
American Heart Association. 2007. Heart disease and stroke statistics—2007 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation (115):e69-e171. http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.106.179918
CDC (Centers for Disease Control and Prevention). 2011. CDC Wide-ranging OnLine Data for Epidemiologic Research (WONDER). Compressed mortality file, underlying cause of death. 1999-2007 (with ICD 10 codes) and 1979-1998 (with ICD 9 codes). Accessed April 2011. http://wonder.cdc.gov/mortSQL.html
CDC. 2004. The burden of chronic diseases and their risk factors—national and state perspectives. http://www.cdc.gov/nccdphp/burdenbook2004/pdf/burden_book2004.pdf
CDC. 1999. Decline in deaths from heart disease and stroke, United States, 1990-1999. Washington, DC.
CDC. n.d. CDC WONDER: Help page for compressed mortality file. Accessed September 2007. http://wonder.cdc.gov/wonder/help/cmf.html
NCHS (National Center for Health Statistics). 2010. Summary health statistics for U.S. adults: National Health Interview Survey, 2009. Vital health Stat. 10(249). http://www.cdc.gov/nchs/data/series/sr_10/sr10_249.pdf
NCHS. 2009a. Summary health statistics for U.S. adults: National Health Interview Survey, 2008. Vital Health Stat. 10(242). http://www.cdc.gov/nchs/data/series/sr_10/sr10_242.pdf
NCHS. 2009b. Summary health statistics for U.S. adults: National Health Interview Survey, 2007. Vital Health Stat. 10(240). http://www.cdc.gov/nchs/data/series/sr_10/sr10_240.pdf
NCHS. 2007. Summary health statistics for U.S. adults: National Health Interview Survey, 2006. Vital Health Stat. 10(235). http://www.cdc.gov/nchs/data/series/sr_10/sr10_235.pdf
NCHS. 2006a. Summary health statistics for U.S. adults: National Health Interview Survey, 2005. Vital Health Stat. 10(232). http://www.cdc.gov/nchs/data/series/sr_10/sr10_232.pdf
NCHS. 2006b. Summary health statistics for U.S. adults: National Health Interview Survey, 2004. Vital Health Stat. 10(228). http://www.cdc.gov/nchs/data/series/sr_10/sr10_228.pdf
NCHS. 2005. Summary health statistics for U.S. adults: National Health Interview Survey, 2003. Vital Health Stat. 10(225). http://www.cdc.gov/nchs/data/series/sr_10/sr10_225.pdf
NCHS. 2004. Summary health statistics for U.S. adults: National Health Interview Survey, 2002. Vital Health Stat. 10(222). http://www.cdc.gov/nchs/data/series/sr_10/sr10_222.pdf
NCHS. 2003. Summary health statistics for U.S. adults: National Health Interview Survey, 2001. Vital Health Stat. 10(218). http://www.cdc.gov/nchs/data/series/sr_10/sr10_218.pdf
NCHS. 2002. Summary health statistics for U.S. adults: National Health Interview Survey, 2000. Vital Health Stat. 10(215). http://www.cdc.gov/nchs/data/series/sr_10/sr10_215.pdf
NCHS. 2001. Summary health statistics for U.S. adults: National Health Interview Survey, 1999. Vital Health Stat. 10(212). http://www.cdc.gov/nchs/data/series/sr_10/sr10_212.pdf
NCHS. 2000. Summary health statistics for U.S. adults: National Health Interview Survey, 1998. Vital Health Stat. 10(209). http://www.cdc.gov/nchs/data/series/sr_10/sr10_209.pdf
NCHS. 1999. Summary health statistics for U.S. adults: National Health Interview Survey, 1997. Vital Health Stat. 10(205). http://www.cdc.gov/nchs/data/series/sr_10/sr10_205.pdf
NIH (National Institutes of Health). 2004. NIH news: The increasing number of adults with high blood pressure. http://www.nhlbi.nih.gov/new/press/04-08-23.htm
State of California. 2005. Proposed identification of environmental tobacco smoke as a toxic air contaminant. Part B: Health effects assessment for environmental tobacco smoke. As approved by the Scientific Review Panel on June 24, 2005. California Environmental Protection Agency, Office of Environmental Health Hazard Assessment. http://www.oehha.org/air/environmental_tobacco/pdf/app3partb2005.pdf
U.S. EPA (United States Environmental Protection Agency). 2004. Air quality criteria for particulate matter. Volumes I (EPA/600/P-99/002aF) and II (EPA/600/P-99/002bF). National Center for Environmental Assessment—RTP Office, Office of Research and Development.
|Cardiovascular Disease Prevalence and Mortality|
|2.||ROE Question(s) This Indicator Helps to Answer|
|This indicator is used to help answer one ROE question: "What are the trends in human disease and conditions for which environmental contaminants may be a risk factors including across population subgroups and geographic regions?"|
This indicator describes data on cardiovascular disease (CVD) prevalence and deaths across the U.S. for the time periods 1997–2009 and 1979–2007, respectively. Cardiovascular disease, which is the leading cause of death and disability in the U.S., may be partly caused or exacerbated by environmental exposures. This information helps to show how rates are changing over time, space, and subpopulations.
The CVD prevalence indicator is based on data published by the Centers for Disease Control and Prevention’s (CDC’s) National Center for Health Statistics (NCHS) that report the results of a series of survey questions pertaining to selected circulatory diseases as part of its National Health Interview Study (NHIS) for the period 1997–2009.
The CVD mortality indicator is based on data archived in the National Vital Statistics System (NVSS) for the period 1979 through 2007.
CVD prevalence data were obtained from annual reports published by CDC’s NCHS, which summarize health statistics compiled from the center’s NHIS (http://www.cdc.gov/nchs/nhis/nhis_series.htm). To maintain respondent confidentiality in NHIS, NCHS withholds variables from the NHIS public use data files that could permit explicit or implicit identification of survey respondents. One of the major risks for inadvertent respondent identification is the inclusion of identifiers on survey files that place respondents in small geographic areas (e.g., census block or census block group). Thus, in these data releases, variables identifying specific geographic areas smaller than one of the four census regions usually are withheld to protect respondent confidentiality (Botman et al., 2000).
CVD mortality statistics were obtained from CDC’s "compressed mortality" database, accessed through CDC’s online database CDC WONDER (Wide-ranging Online Data for Epidemiologic Research), at http://wonder.cdc.gov/mortSQL.html. The number of deaths, crude death rates, or age-adjusted death rates can be obtained for total U.S., state, or county; by age group; race (white, black, and other); gender; year of death; and underlying cause–of–death (ICD code). Individual-level data are not available due to confidentiality issues.
The NHIS is the nation’s primary source of general health information for the resident civilian non-institutionalized population of the United States and since 1960 has been one of the major data collection programs of the NCHS. The NHIS data for the reporting period (1997–2009) have been designed to produce reliable annual national estimates. The NHIS section on health conditions utilizes survey questions specifically designed to produce national estimates of disease prevalence. CDC collects cardiovascular disease prevalence data via a series of survey questions pertaining to selected circulatory diseases (i.e., “coronary heart disease [CHD],” “stroke,” “hypertension,” and “all types” of CVD as part of its NHIS.
The NHIS is a cross–sectional household survey, with sampling and interviewing conducted continuously throughout each year. The survey is designed so the sample scheduled for each week is representative of the target population, and the weekly samples are additive over time. The sampling plan follows a multistage area probability design that permits the representative sampling of households. The sampling plan was redesigned in 1995 to include the over sampling of certain populations to ensure adequate representation (e.g., black and Hispanic persons), and to draw samples from each state. Although the NHIS sample is too small to provide state level data with acceptable precision for each state, this design facilitates the collection of NHIS data through the use of state-level telephone health surveys.
These data represent a comprehensive and reliable data set for measuring trends in CVD prevalence across the U.S (nationwide and regionally), over time, and across subgroups (Botman et al., 2000). The NHIS questionnaire was revised extensively in 1997, which marks the beginning of the reporting period for this indicator. The revised NHIS questionnaire has “core” questions and “supplements.” The core questions are largely unchanged from year to year and allow for trends analysis and for data from more than one year to be pooled to increase sample size for analytic purposes. The core questions contain four major components: household, family, sample adult, and sample child. The NHIS CVD questions are included within the sample adult component.
The response rate for the ongoing portion of the survey (core) has been between 94 and 98 percent over the years. Response rates for special health topics (supplements) have generally been lower.
Standard documentation is available to support the underlying prevalence data. Complete program methodology (including a description of the survey and the survey methodology) can be found at:
A detailed description of the underlying sampling procedures for the NHIS can be found in Botman et al. (2000), which is applicable to NHIS data collected from 1999 to 2009.
The NVSS is responsible for collecting the nation’s official vital statistics. These vital statistics are provided through state-operated registration systems. Virtually all deaths nationwide are registered with the NVSS. Data are collected from all 50 states and the District of Columbia. The number of deaths attributed to birth defects can be obtained for total U.S., state, or county; by age group; race (white, black, and other); gender; year of death; and underlying cause of death (ICD code).
Standard forms for the collection of data and model procedures for the uniform registration of the events are developed and recommended for state use through cooperative activities of the states and the NCHS. U.S. standard death certificates (see http://www.cdc.gov/nchs/data/dvs/DEATH11-03final-ACC.pdf (4 pp, 556K)) are revised periodically. Most state certificates conform closely in content and arrangement to the standard certificate recommended by NCHS and all certificates contain a minimum data set specified by NCHS. Demographic information on the death certificate is provided by the funeral director based on information supplied by an informant. A physician, medical examiner, or coroner provides medical certification of cause of death.
General information regarding data collection procedures can be found in the Model State Vital Statistics Act and Regulations (CDC, 1995). Documentation is also available at http://wonder.cdc.gov/wonder/help/cmf.html.
The recommended certificate of death is posted at http://www.cdc.gov/nchs/data/dvs/DEATH11-03final-ACC.pdf (4 pp, 556K) and the documentation for the mortality data set is at http://wonder.cdc.gov/wonder/help/cmf.html.
Prevalence data are reported as frequencies and percentages of the U.S. population as reported in NHIS summary health statistics reports. CVD prevalence indicators are expressed as crude rates for the number of reported cases per 1,000 people. Crude rates were used to maintain consistency of methodology throughout the reporting period because CDC did not report age-adjusted data prior to 2002 in the data sources used for this indicator.
CVD mortality data are reported as age-adjusted deaths per 100,000 people. Mortality data are queried from CDC’s "compressed mortality" database, accessed through CDC WONDER (CDC, 2011). The international classification of disease (ICD) ninth and tenth revision codes were used to select the specific type of CVD (e.g., CHD or stroke) (see http://wonder.cdc.gov/wonder/help/cmf.html).
This indicator uses the following ICD-9 cause of death classifications:
This indicator uses the following ICD-10 cause of death classifications:
For national (i.e., all states combined) data, all rates are queried using the 2000 U.S. standard population age-adjusted rate option expressed as number of cases per 100,000 population. Beginning with the 1999 data year, NCHS adopted the year 2000 projected population of the United States as the standard population. These age-adjusted rates are used without additional statistical manipulation for reporting national mortality data for CVD (all), CHD, hypertension, myocardial infarction (MI), and stroke.
For EPA Regional breakdowns, indicator data were generated by calculating state-specific age-adjusted mortality rates for the years 1979 through 2007. Crude (i.e., not age-adjusted) CVD mortality rates obtained from CDC WONDER were tabulated by state and age group. For each age group and state, the number of cases reported and the corresponding population for the age cohort were used to calculate age–adjusted rates. The state–by–state totals stratified by the specified age groups were combined for each EPA Region. Rates were age–adjusted using the 2000 U.S. standard population.
Age-adjusted mortality rates are widely used in epidemiologic and disease surveillance practices and are an accepted indicator of the disease burden attributed to CVD–related illness across the United States. Age–adjusted mortality rates are routinely used to compare mortality rates across groups and over time.
|9.||Quality Assurance and Quality Control|
NHIS data collection is conducted by the U.S. Bureau of the Census under an interagency agreement with NCHS. Interviewers receive extensive training, and their work is monitored through a quality assurance program. Quality assurance and quality control (QA/QC) information is provided in Botman et al. (2000).
Vital statistics regulations have been developed to serve as a detailed guide to state and local registration officials who administer the U.S. vital statistics system. These regulations provide specific instructions to protect the integrity and quality of the data collected. The quality assurance information can be found in CDC (1995).
This indicator does not have established reference points or thresholds.
|11.||Comparability Over Time and Space|
Because a single interviewing staff work year-round to collect NHIS data from across the nation, this indicator should be fairly comparable spatially and temporally. However, under-diagnosis rates vary by demographic and have changed over time.
Prevalence data are generally comparable over the time period covered by this indicator. However, because the NHIS questionnaire was redesigned in 1997, data from 1997 and later years may not be comparable with earlier years. For example, in 1997, the collection methodology changed from paper and pencil questionnaires to computer-assisted personal interviewing (CAPI). Also, in some instances basic concepts measured in NHIS changed, while in other cases, concepts remained the same but were measured differently. In addition, some questions remained the same over time, but were preceded by different questions or topics. Further, for some questions, there was a change in the reference period for reporting an event or condition.
Due to differences in the ICD system used for classifying mortality, data from 1979 to 1998 (ICD–9 codes) cannot be directly compared with data from 1999 to 2007 (ICD-10 codes) (also see "Indicator Derivation").
Prevalence and mortality data are generally comparable over space, as both are considered nationally representative. Data collection is standardized and generally consistent across the nation.
|12.||Sources of Uncertainty|
Content under review.
|13.||Sources of Variability|
Content under review.
No trend analysis has been conducted on this dataset.
Limitations to this indicator include the following:
Botman, S.L., T.F. Moore, C.L. Moriarity, and V.L. Parsons. 2000. Design and estimation for the National Health Interview Survey, 1995–2004. National Center for Health Statistics. Vital Health Stat. 2(130). http://www.cdc.gov/nchs/data/series/sr_02/sr02_130.pdf (41 pp, 300K).
CDC (Centers for Disease Control and Prevention). 2011. CDC Wide-ranging OnLine Data for Epidemiologic Research (WONDER). Compressed mortality file, underlying cause of death. 1999-2007 (with ICD 10 codes) and 1979-1998 (with ICD 9 codes). Accessed April 2011. http://wonder.cdc.gov/mortSQL.html.
CDC. 1995. Model State Vital Statistics Act and Regulations (revised April 1995). DHHS publication no. (PHS) 95-1115. http://www.cdc.gov/nchs/data/misc/mvsact92aacc.pdf (38 pp, 2.9MB).
NCHS (National Center for Health Statistics). 2007. Summary health statistics for U.S. adults: National Health Interview Survey, 2006. Vital Health Stat. 10(234). http://www.cdc.gov/nchs/data/series/sr_10/sr10_235.pdf (163 pp, 3.6MB).