Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Oct 1.
Published in final edited form as: Soc Sci Med. 2017 Aug 8;190:174–180. doi: 10.1016/j.socscimed.2017.08.004

Why did life expectancy decline in the United States in 2015? A gender-specific analysis

Francesco Acciai *, Glenn Firebaugh *
PMCID: PMC5611857  NIHMSID: NIHMS903184  PMID: 28865253

Abstract

In 2015, age-adjusted mortality rates increased for 8 of the 10 leading causes of death in the United States. As a result, life expectancy declined by 0.17 years for both women and men. The decline could be just an anomaly, or it could represent the start of a new trend of stagnation or decline in life expectancy, as some scholars have warned. The first step is to determine the sources of the decline. In this study we analyze the contribution of specific causes of death to the decline in men’s versus women’s life expectancy, as well as the mechanisms that generated the decline. With regard to mechanisms, we examine whether the decline in life expectancy resulted from reductions in the average age at death for most causes of death, or from a greater risk of dying of causes that disproportionately affect the young and middle-aged. Using CDC mortality data, we construct life tables for men and women separately, based on 20 different underlying causes of death. We find that men’s reduction in life expectancy was attributable primarily to changes in midlife mortality, with a greater risk of dying of causes such as accidental poisoning or homicide. The reduction in women’s life expectancy, on the other hand, was attributable primarily to changes in old-age mortality, where the mean age at death for most causes (such as heart disease and mental illnesses) was lower in 2015 than it was in 2014. These gender-specific mechanisms that contributed to the decline in life expectancy are common to the major racial/ethnic groups, but the magnitude of the decline differs greatly across racial/ethnic groups. Future research can use the tools provided here to investigate in more detail how the gender-specific mechanisms of the 2015 decline differ by race.

Keywords: Life expectancy, gender, causes of death, mortality, age-risk decomposition, accidental poisoning

INTRODUCTION

In 2015 life expectancy at birth (e0) in the United States was lower than it was in 2014 (Xu et al., 2015). The previous reductions in life expectancy observed in the United States date back to 1993, during the HIV/AIDS epidemic, and to 1980, a year characterized by a particularly deadly influenza (OECD, 2017; The Associated Press, 2016). The decrease in life expectancy observed in 2015 is particularly worrisome for two reasons. First, unlike the declines in 1980 and 1993, the 2015 decline in life expectancy was not generated by an anomalous spike in a specific cause of death. Instead, the age-adjusted death rates increased for 8 of the 10 leading causes of death—heart disease, chronic lower respiratory diseases, accidents, stroke, Alzheimer’s, diabetes, kidney disease, and suicide (Xu et al., 2015). Second, the decrease in life expectancy in 2015 occurred after several years of stagnation in the life expectancy trend, as life expectancy rose very little from 2010 to 2014 (NCHS, 2016; OECD 2017).

The five year long period of stagnation in life expectancy followed by an actual decline represents a clear discontinuity in the secular trend of life expectancy in the United States. Indeed, life expectancy in the United States increased throughout the 20th century and the first decade of the 21st century (Kinsella, 1992). Men and women born in 1900 could expect to live on average 48.3 and 51.1 years, respectively. By 1950 life expectancy had risen by over 20 years for women (e0=71.7) and by almost 18 years for men (e0=66). This unprecedented increase in life expectancy derived from a substantial decrease in the mortality rate of infectious diseases, fostered by improved nutrition (McKeown, 1976), population-wide public health actions (e.g. water sanitation, vaccination campaigns, introduction of antibiotics), and increasingly widespread socioeconomic well-being (Cutler et al., 2006; Riley, 2001). These factors significantly lowered infant and young-age mortality (Cutler et al., 2006; Oeppen and Vaupel, 2002). On the other hand, in recent decades the rise in life expectancy has been generated mostly by lower mortality rates at older ages (i.e. 65+) (Cutler et al., 2006; Oeppen and Vaupel, 2002; Riley, 2001), thanks to medical improvements in the detection and treatment of chronic and degenerative diseases—the most common causes of death.

In the current paper we focus on the 2015 decline in life expectancy by analyzing the changes in the mortality regime of the two major subgroups of the population—men and women. This distinction is important because men and women differ significantly in terms of life expectancy as well as in cause- and-age-specific mortality rates. Therefore, overall trends might mask gender-specific patterns. The overarching goal of the paper is to compare the drivers of the 2015 decline in men’s and women’s life expectancy. First we compare the contributions of the most common and relevant causes of death to the decline in men’s versus women’s life expectancy. Then we investigate the underlying mechanisms that generated the decline for women versus men by distinguishing between changes in cause-specific mean age at death and changes in the cause-specific probability of dying. In other words, we address the following question: Did life expectancy for women (men) decline because of a greater risk of dying of causes that disproportionately affect the young (lifetime risk component), or because women (men) are dying of the same causes as before, but at a younger age (age component)?

MATERIAL AND METHODS

Data

We used the mortality data and population denominators from the Multiple Cause of Death Mortality files of the National Center for Health Statistics (NCHS, 2016) to construct gender-specific multiple-decrement life tables for 2014 and 2015. Our life tables are based on 5-year age groups (but age 0–4 is separated into 0–1 and 1–4) until age 85, plus an open age group (85+). We include virtually the entire U. S. population; the only omissions are non-Hispanic Native Americans and Alaska Natives, and individuals whose Hispanic origin or age at death is unknown.

Methods

Our classification of causes of death is based on the tenth revision of the International Classification of Diseases (ICD10) (WHO, 2008). We started with the ten leading causes of death as defined by the CDC (Centers for Disease Control and Prevention) (Heron, 2016), to which we added other relevant causes (septicemia; chronic liver disease and cirrhosis; HIV/AIDS; homicide; infant conditions; mental and behavioral disorders; diseases of the digestive system; ill-defined causes) and a residual category (minor causes not elsewhere classified). In addition, we split “accidents” – one of the CDC’s ten leading causes – into three separate causes (accidental poisoning; transport accidents; falls). Death due to accidental poisoning has been increasing at an alarming rate over the past 15 years (Case and Deaton, 2015), reducing the growth in life expectancy observed between 2000 and 2014, for men particularly. Therefore, we suspect that accidental poisoning could be a key driver of the 2015 decline in life expectancy, especially for men. We also distinguished between falls and traffic accidents because these causes have different age and gender profiles. Lastly, we moved Alzheimer’s into a larger cause category labeled “mental illnesses”, which also includes all other mental and behavioral disorders. Overall, we distinguish among 19 different causes of death plus a residual category (see Appendix Table 1 for the specific ICD10 coding).

When multiple causes of death are involved, life expectancy at birth (e0) is the weighted average of the cause-specific mean ages at death, where the weights are the cause-specific lifetime risks. In symbols, e0= Σkakrk, where a and r indicate the mean age death and lifetime risk, respectively, and the subscript k (k=1,…,20) refers to the cause of death. The lifetime risks are, by definition, the probability of dying of each cause, derived from the period life table. Therefore, Σ krk= 1 when, as in the current study, the causes of death are exhaustive and mutually exclusive.

Any change in the cause-specific mean ages at death or in the cause-specific probabilities of dying will alter the overall life expectancy. These two processes usually operate simultaneously and, as previous work (Acciai et al., 2015; Firebaugh et al., 2014) has shown, it is possible to separate and quantify their contribution to the total change in life expectancy:

 2015re0 2014re0=AgeComponent+RiskComponent.

Distinguishing between these two components provides useful insights into how the mortality regime has changed over time. Specifically, we can quantify the proportion of the decline in life expectancy that was generated by lower mean ages at death (age component) and the proportion generated by a change in the cause-specific probabilities of dying—with a reallocation of deaths toward younger-age causes (lifetime risk component). Furthermore, it is possible to make the distinction between age and risk component for each cause of death separately (Acciai et al., 2015; Firebaugh et al., 2014). The total age and risk components are, in fact, the sum of the cause-specific age and risk components: Age Component = Σk Agek; Risk Component = Σk. Riskk.

RESULTS

Descriptive Results

Table 1 reports the cause-specific lifetime risks and average ages at death from the life tables for 2014 and 2015, for both men and women. The total mean age at death (last row of the table) corresponds to life expectancy at birth. Importantly, men and women experienced a decline of the same magnitude—for both groups life expectancy in 2015 was 0.17 years lower than it was in 2014. The descriptive statistics can help determine why this occurred. For instance, the risk of dying of most of the midlife causes, such as traffic accidents or accidental poisoning, is greater in 2015. This pattern is observed for both genders, but was more marked for men. Another element that depressed life expectancy in 2015 was the decline in the average age at death of the most common—and therefore most relevant—causes, such as heart disease and cancers. Since these two causes combined account for over 47% of all deaths for men and over 42% of all deaths for women, even a small reduction in their mean age at death can have a noticeable negative effect on life expectancy.

Table 1.

Gender-specific proportions (lifetime risks) and mean ages at death, by cause of death: US 2014–2015.

Cause of death MEN WOMEN

Proportion (risk) Mean age at death Proportion (risk) Mean age at death

2014 2015 2014 2015 2014 2015 2014 2015
Chronic and degenerative diseases
 Heart diseases 0.2592 0.2582 80.02 79.96 0.2378 0.2373 85.19 84.98
 Cancers 0.2231 0.2176 76.18 76.13 0.1910 0.1868 76.46 76.37
 Mental illnesses 0.0806 0.0782 86.12 85.92 0.1483 0.1442 89.26 89.05
 Chronic lower respiratory diseases 0.0566 0.0569 80.39 80.35 0.0596 0.0616 81.61 81.65
 Cerebrovascular diseases 0.0465 0.0474 81.33 81.43 0.0648 0.0663 85.52 85.44
 Diabetes 0.0291 0.0294 75.90 75.72 0.0255 0.0253 79.44 78.96
 Diseases of the digestive system 0.0216 0.0213 77.34 77.11 0.0256 0.0255 81.68 81.44
 Nephritis, nephrotic syndrome and nephrosis 0.0038 0.0036 81.11 80.87 0.0037 0.0037 83.66 83.20
Communicable diseases
 Influenza and pneumonia 0.0226 0.0228 81.78 82.65 0.0231 0.0242 84.26 85.21
 Septicemia 0.0140 0.0144 78.11 78.29 0.0153 0.0154 80.39 80.31
 HIV/AIDS 0.0023 0.0022 53.20 54.18 0.0008 0.0008 50.79 51.50
External causes
 Suicide 0.0169 0.0170 52.61 52.05 0.0046 0.0049 49.04 48.82
 Alcoholic liver diseases and cirrhosis 0.0129 0.0132 63.92 63.89 0.0074 0.0078 66.64 66.42
 Accidental poisoning 0.0126 0.0144 43.01 42.62 0.0073 0.0078 46.44 45.59
 Transport accident 0.0133 0.0139 48.71 48.64 0.0055 0.0058 51.76 51.02
 Falls 0.0137 0.0140 81.84 81.83 0.0136 0.0137 86.63 86.61
 Homicide 0.0055 0.0063 34.90 34.39 0.0016 0.0017 40.60 40.62
Other causes
 Infant conditions 0.0058 0.0058 11.69 12.95 0.0051 0.0051 13.50 13.74
 Ill-defined causes 0.0107 0.0104 71.51 69.86 0.0140 0.0131 81.00 80.00
 NEC - Not elsewhere classified 0.1492 0.1530 77.85 77.89 0.1454 0.1493 81.30 81.26

Total 1.0000 1.0000 76.926 76.752 1.0000 1.0000 81.760 81.586

Cause-specific analysis

Cause-specific analysis reveals that some of the major contributors to the decline in e0 are common to men and women. Figures 1A and 1B summarize the causes that contributed to lowering life expectancy in 2015 by at least 0.01 years (or over 5% of the decline) for women and for men, respectively. (Table 3, below, reports results for all 20 causes.) Except for diabetes (only relevant for women) and homicide and suicide (only relevant for men), the other five major contributors to the decline in e0 are the same for both genders. These are heart diseases, transport accident, accidental poisoning, mental illnesses, and ill-defined causes. However, these causes contribute to the decline in life expectancy to a different extent for men and women. For women the major contributors are heart disease and mental illness, followed by accidental poisoning. For men the major contributors are accidental poisoning, mental illnesses, and homicide, in that order. Overall, women’s life expectancy decreased mostly because of old-age causes of death, whereas men’s life expectancy was more affected by midlife causes of death, with the exception of mental illnesses.

Figure 1A.

Figure 1A

Causes that significantly contributed to the 2015 decline in women’s life expectancy

Figure 1B.

Figure 1B

Causes that significantly contributed to the 2015 decline in men’s life expectancy

Note: The longer the bar, the greater the reduction in life expectancy attributable to the cause.

Table 3.

Cause-specific age-risk decomposition results. US 2014–2015, men and women.

Cause of death MEN WOMEN

Risk Age TOTAL Risk Age TOTAL
Chronic and degenerative diseases
 Heart diseases −0.003 −0.014 −0.017 −0.002 −0.050 −0.052
 Cancers 0.004 −0.010 −0.006 0.023 −0.018 0.005
 Mental illnesses −0.022 −0.016 −0.038 −0.030 −0.031 −0.061
 Chronic lower respiratory diseases 0.001 −0.002 −0.001 0.000 0.002 0.002
 Cerebrovascular diseases 0.004 0.005 0.009 0.006 −0.005 0.001
 Diabetes 0.000 −0.005 −0.005 0.001 −0.012 −0.012
 Diseases of the digestive system 0.000 −0.005 −0.005 0.000 −0.006 −0.006
 Nephritis, nephrotic syndrome and nephrosis −0.001 −0.001 −0.002 0.000 −0.002 −0.002
Communicable diseases
 Influenza and pneumonia 0.001 0.020 0.020 0.003 0.022 0.026
 Septicemia 0.001 0.002 0.003 0.000 −0.001 −0.001
 HIV/AIDS 0.002 0.002 0.004 0.002 0.001 0.002
External causes
 Suicide −0.003 −0.009 −0.012 −0.007 −0.001 −0.008
 Alcoholic liver diseases and cirrhosis −0.004 0.000 −0.004 −0.007 −0.002 −0.009
 Accidental poisoning −0.061 −0.005 −0.066 −0.018 −0.006 −0.025
 Transport accident −0.020 −0.001 −0.020 −0.007 −0.004 −0.011
 Falls 0.002 0.000 0.001 0.000 0.000 0.000
 Homicide −0.031 −0.003 −0.034 −0.003 0.000 −0.003
Other causes
 Infant conditions −0.002 0.007 0.005 −0.001 0.001 0.001
 Ill-defined causes 0.002 −0.017 −0.015 0.001 −0.014 −0.012
 NEC - Not elsewhere classified 0.004 0.005 0.009 −0.002 −0.007 −0.009

TOTAL 0.127 0.048 0.175 0.042 0.133 0.174

Midlife mortality vs old-age mortality

To further explore whether, and how, women and men differed in the mortality patterns that led to a reduction in life expectancy, we determine whether the 2015 reduction in life expectancy at birth (e0) mostly occurred before or after age 65 (Table 2). Since Δe0 = Δe064 + Δe65 we can easily determine what portion of the change in e0 is attributable to changes in old age mortality, captured by change in life expectancy at age (e65). Although the decline in e0 was the same for women and men, the mechanisms were quite different. In fact, changes in old-age mortality generated most of the decline (0.136/0.174, or 78%) for women, but only 33% (0.057/0.175) of the decline for men. This pattern is in line with the findings of our cause-specific analysis, where we found that the major contributors of the decline were primarily old-age causes of death for women, and midlife causes of death for men.

Table 2.

Life expectancy at birth and at age 65 for men and women

Life expectancy at birth (e0) Life expectancy at 65 (e65)

2014 2015 Change 2014 2015 Change
Men 76.926 76.752 −0.175 18.529 18.472 −0.057
Women 81.760 81.586 −0.174 21.145 21.009 −0.136

Age-Risk decomposition results

Although the principal causes of death (e.g. heart disease, mental illnesses, accidental poisoning) responsible for the decline in life expectancy were largely the same for men and women, the mechanisms of the decline were very different. The age-risk decompositions show that the risk component drove the decline for men while the age component drove the decline for women (Table 3, last row). In other words, men’s life expectancy decreased largely because younger-age causes were more common in 2015 than they were in 2014, whereas women’s life expectancy fell largely because women tended to die of the same causes but at younger ages in 2015 than in 2014.

Detailed results of the cause-specific age-risk decomposition are reported in Table 3. For each cause the decomposition provides the age and the risk components, the sum of which corresponds to the cause-specific total contribution (in years) to the decline in life expectancy. Notably, only one cause—influenza and pneumonia—significantly offset the decline in life expectancy, with a non-trivial positive contribution for both men and women. One might have expected that cancer—a cause that accounts for about one in five deaths in the U.S.—would also have offset the decline in life expectancy, since the age-adjusted death rate for cancer decreased in 2015 (Xu et al., 2015). However, as Table 3 shows, the decline in cancer death rates—captured by a positive risk component—is offset by the fact that cancer victims tended to be somewhat younger in 2015 (see Table 1), as shown by the negative age component.

Finally, we note that our results closely align with results for whites, who represent the largest share of the population. However, race-specific analyses reveal that, for every racial/ethnic group, the risk component dominated for men while the age component dominated for women. The magnitudes of the age and risk components nonetheless varied across the groups (Figures 2A and 2B). The risk component was particularly striking for black men: The greater risk of dying of midlife causes was responsible for reducing blacks’ life expectancy by 0.28 years (Figure 2B), nearly half of which was due to homicide (cause-specific results for racial groups are not shown here, but are available upon request). In contrast, the greater risk of dying of midlife causes reduced life expectancy by a little over 0.10 years for white men and Hispanic men, with accidental poisoning being the main culprit. In the case of women, the age component was largest for white women (Figure 2A), due mainly to reductions in the ages of white women who died of older-age causes such as heart disease or mental disorder. While the age component is much smaller for black and Hispanic women, it still accounts for most of their decline in life expectancy.

Figure 2A.

Figure 2A

Age-risk decomposition for women by racial/ethnic group

Figure 2B.

Figure 2B

Age-risk decomposition for men by racial/ethnic group

DISCUSSION

2015 marked the first decline in life expectancy in the United States in over two decades. Despite an historical trend of rising life expectancy in the U.S., some scholars had in fact warned of the possibility that this trend might soon be reversed. For instance, over a decade ago Olshansky and colleagues (2005) argued that if the prevalence of obesity continues to increase, life expectancy will be negatively affected at some point during the first half of the 21st century. Their forecast was corroborated by more recent work (Ludwig, 2016; Stewart et al., 2009). But even though the death rates for obesity-related causes were higher in 2015 than they were in 2014 (Ludwig, 2016) the decline in life expectancy observed in 2015 does not seem to be closely related to obesity, because (1) most causes of death contributed to the decline, and (2) most of the decline for men was attributable to external causes of death.

In fact, obesity is not the only risk factor that has raised concerns. Socioeconomic hardship and instability is considered one of the drivers of the deaths of despair—a group of external causes that includes accidental poisoning, suicide, and chronic liver diseases (Case and Deaton, 2015; Masters et al., 2017). For all these causes, though to different extents, the death rates have increased over the past 15 years, with white males aged 45–54 being especially affected (Case and Deaton, 2015). This trend is particularly worrisome because even small increases in the age-adjusted death rates for midlife causes can have a non-trivial impact on life expectancy. Extending the findings of Case and Deaton (2015), we find that the rise in accidental poisoning has significantly depressed life expectancy for all Americans, and not just for white men. This is even more striking if we consider that accidental poisoning accounts for only about 1.3% of men’s deaths and 0.75% of women’s death. The other deaths of despair, suicide and alcoholic liver diseases, also contributed to lowering life expectancy in 2015 (Table 3), but with the exception of suicide for men, their contribution was only marginal. However, other external causes such as homicide (especially for black men) and transport accidents also contributed to the decline in life expectancy.

To summarize, the increase in mortality that produced the 2015 decline in life expectancy was not limited to one or two causes of death. To the contrary, the mean age at death decreased for most causes of death (14 of 20 causes for men, and 15 of 20 causes for women), resulting in a large negative age component, especially for white women. For men the negative age component was compounded by an even larger negative risk component fueled by rising rates of accidental poisoning among whites and Hispanics, and homicide among blacks. To gain further insights on the trends and mechanisms of the 2015 decline in life expectancy, future studies should examine race- and gender-specific patterns in more detail. The methods we use here could also be employed to study regional patterns as well as to determine whether the decline has affected individuals from different socioeconomic backgrounds differently.

RESEARCH HIGHLIGHTS.

  • 2015 marked the first decline in life expectancy in the United States since 1993

  • Men and women experienced a decline in life expectancy of similar magnitude

  • Women’s decline was driven mostly by an increase in old-age mortality (age 65+)

  • Men’s decline was driven mostly by an increase in midlife mortality

  • Key drivers are accidental poisoning for men and mental illnesses for women

Acknowledgments

This research was supported by Grant R03AG050895-02 from the National Institutes of Health as well as by funding from NICHHD Grant R24HD041025 to the Population Research Institute (PRI) at the Pennsylvania State University. The funding sources had no role in the design and conduct of the study. We are very grateful to Steven Haas for his insightful comments and encouragement.

APPENDIX

Appendix Table 1.

ICD10 coding scheme utilized to create the relevant cause categories

Cause of death ICD10 code

Chronic and degenerative diseases
 Heart diseases I00–I09, I11, I13, I20–I51
 Cancers C00–C97
 Mental illnesses G30, F00–F99
 Chronic lower respiratory diseases J40–J47
 Cerebrovascular diseases I60–I69
 Diabetes E10–E14
 Diseases of the digestive system K00–K69, K71–72, K75–K93
 Nephritis, nephrotic syndrome and nephrosis N00–N07, N17–N19, N25–N27
Communicable diseases
 Influenza and pneumonia J09–J18
 Septicemia A40–A41
 HIV/AIDS B20–24
External causes
 Suicide X60–X84, Y87, U03
 Alcoholic liver diseases and cirrhosis K70, K73–K74
 Accidental poisoning X40–X45, Y10–Y19, Y45, Y47, Y49, T36–T50, T96
 Transport accident V00–V99
 Falls W00–W19
 Homicide X85–X99, Y00–Y09, U01–U02
Other causes
 Infant conditions P00–P96, Q00–Q99
 Ill-defined causes R00–R99
 NEC - Not elsewhere classified All remaining causes

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Acciai F, Noah AJ, Firebaugh G. Pinpointing the sources of the Asian mortality advantage in the United States. J Epidemiol Commun H. 2015;69(10):1006–1011. doi: 10.1136/jech-2015-205623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Case A, Deaton A. Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century. P Natl Acad Sci USA. 2015;112(49):15078–15083. doi: 10.1073/pnas.1518393112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cutler D, Deaton A, Lleras-Muney A. The Determinants of Mortality. Journal Econ Perspect. 2006;20(3):97–120. [Google Scholar]
  4. Firebaugh G, Acciai F, Noah AJ, Prather CJ, Nau C. Why the racial gap in life expectancy is declining in the United States. Demographic Res. 2014;31(32):975–1006. doi: 10.4054/DemRes.2014.31.32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Heron M. Deaths: Leading causes of death for 2014. National Center for Health Statistics (NCHS) National Vital Statistics Report. 2016;65(5) [PubMed] [Google Scholar]
  6. Kinsella KG. Changes in life expectancy 1900–1990. Am J Clin Nutr. 1992;55:1196S–1202S. doi: 10.1093/ajcn/55.6.1196S. [DOI] [PubMed] [Google Scholar]
  7. Ludwig DS. Lifespan weighted down by diet. JAMA. 2016;315(21):2269–2270. doi: 10.1001/jama.2016.3829. [DOI] [PubMed] [Google Scholar]
  8. Masters RK, Tilstra AM, Simon DH. Mortality from Suicide, Chronic Liver Disease, and Drug Poisonings among Middle-Aged U.S. White Men and Women 1980–2013. Biodemography Soc Biol. 2017;63(1):31–37. doi: 10.1080/19485565.2016.1248892. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. McKeown T. The Modern Rise of Population. New York, NY: Academic Press; 1976. [Google Scholar]
  10. National Center for Health Statistics (NCHS) Health, United States, 2015: With Special Feature on Racial and Ethnic Health Disparities. Hyattsville, MD: 2016. p. 95. Table 15. [PubMed] [Google Scholar]
  11. National Center for Health Statistics (NCHS) Multiple Cause of Death File 2014–2015. Washington, DC: 2016. https://wonder.cdc.gov/mcd.html. [Google Scholar]
  12. OECD. Life expectancy at birth (indicator) 2017 doi: 10.1787/27e0fc9d-en. Accessed on 15 March. 2017. [DOI] [Google Scholar]
  13. Oeppen J, Vaupel JW. Broken limits to life expectancy. Science. 2002;296(5570):1029–1031. doi: 10.1126/science.1069675. [DOI] [PubMed] [Google Scholar]
  14. Olshansky SJ, Passaro DJ, Hershow RC, et al. A potential decline in life expectancy in United States in the 21st century. New Engl J Med. 2005;352(11):1138–1145. doi: 10.1056/NEJMsr043743. [DOI] [PubMed] [Google Scholar]
  15. Riley JC. Rising Life Expectancy: A Global History. New York: Cambridge University Press; 2001. [Google Scholar]
  16. Stewart ST, Cutler DM, Rosen AB. Forecasting the effects of obesity and smoking on U.S. life expectancy. New Engl J Med. 2009;361:2252–2260. doi: 10.1056/NEJMsa0900459. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. The Associated Press. US life expectancy falls; SU professor sees ‘troubling trends’ for poorer people. 2016 Dec 9; ( http://www.syracuse.com/us-news/index.ssf/2016/12/us_life_expectancy_falls_su_professor_sees_troubling_trends_for_poorer_people.html). Accessed on 15 March 2017.
  18. World Health Organization (WHO) ICD-10: International statistical classification of diseases and related health problems. World Health Organization; Geneva, Switzerland: 2008. [Google Scholar]
  19. Xu J, Murphy SL, Kochanek KD, Arias E. Mortality in the United States, 2015. (Data Brief – No. 267).National Center for Health Statistics. 2016 [PubMed] [Google Scholar]

RESOURCES