11

Implications for Policy and Research

This chapter provides background for and develops the policy and research implications of the committee’s analysis and findings. Policy implications encompass how the evidence in this report suggests the need for changes in policy with the potential to curb the recent increase in working-age mortality and/or narrow disparities in working-age mortality in the coming years. The committee stresses the immense challenge of predicting policy impacts in this area of science. As reviewed throughout the report, studies of mortality trends and patterns, especially at the national level, rely almost exclusively on observational data and federal statistics. As a result, causal evidence in this area is limited, and controlled experiments are difficult, if not infeasible. Moreover, as also discussed throughout the report, the key hypothesized influences on working-age mortality patterns and trends are numerous and operate concomitantly at multiple levels. Many of the proposed drivers operate across the life course and/or across decades—in either period or cohort fashion—to influence mortality patterns and trends. This report therefore focuses on an exceptionally complex set of patterns, trends, and explanations for which clear or simple solutions are lacking.

Despite such complexity and the necessary reliance on observational and administrative data, the committee emphasizes the urgency of policy action in the face of a population health crisis that is claiming the lives of working-age adults in the prime of their lives, a crisis that has been further exacerbated by the COVID-19 pandemic. The trends and patterns documented in Chapters 2–4 of the report clearly show that life expectancy in the United States is lower than that in all other similar high-income

democratic countries; that working-age mortality rates increased during the period examined for this study (1990–2017) in the United States but not in those other countries, with pronounced increases for particular causes of death; and that disparities in working-age mortality are either stubbornly wide (e.g., non-Hispanic [NH] Blacks [Blacks] compared with non-Hispanic Whites [Whites]) or widening (e.g., individuals with low versus high levels of education). This report demonstrates further that working-age mortality rates are significantly higher, and rising more rapidly, in some geographic areas of the country than in others.

This report also identifies significant knowledge gaps and data limitations that have critical implications for further research designed to isolate the key drivers and modifiable factors that explain the rise of and disparities in working-age mortality. The need for additional research is pressing. From a historical perspective, the rise in U.S. working-age mortality and recent resulting declines in life expectancy are relatively new phenomena. As this report documents, because the rise in working-age mortality was specific to certain causes of death but with varying patterns by age, sex, race and ethnicity, socioeconomic status, and geography, existing research into these complex and multilayered patterns is sparse, and emerging research attempting to better understand the explanations for these changing patterns is nascent. Therefore, much remains to be learned, and the committee offers a number of research recommendations focused on generating better evidence that can serve as the basis for policy evaluation and refinement.

This chapter first presents a framework that establishes three general content areas in which the committee’s policy and research implications are most relevant: (1) medical science and health care access and delivery, (2) public health, and (3) social and economic policy. Within each category, the chapter provides a brief overview of policies that have resulted in success, failure, or some mix of success and failure in the past. With this experience in mind, the chapter then brings together the policy and research implications of earlier chapters to emphasize how those implications fit within these three content areas. Importantly, this chapter also presents new policy and research implications within the three content areas that cut across all of the previous chapters. As part of this discussion, the chapter also steps back to reflect on the key thematic findings regarding economic change, socioeconomic inequality, and vulnerability that are presented across the various chapters of the report. The chapter concludes with a brief discussion of lessons learned from the COVID-19 pandemic.

For ease of reference, Table 11-1 at the end of the chapter lists all of the recommendations and policy conclusions presented in the report, grouped thematically in the categories of opioids, other drugs, and alcohol; suicide; cardiometabolic diseases; cross-cutting themes; and data needs.

A FRAMEWORK FOR THE CATEGORIZATION OF POLICY AND RESEARCH IMPLICATIONS

This chapter draws on historical insights and this report’s findings to suggest ways in which changes in policy and further research are likely to impact future trends and disparities in U.S. working-age mortality. As noted above, this discussion is organized into three broad categories of potential policy intervention and areas for more research:

• Medical science and health care access and delivery. This category includes advances in medical science, such as new pharmaceuticals and surgeries to treat patients with disease. This category also includes policies that provide working-age individuals with greater access to health care (e.g., Medicaid) and that lead to advances in medical practice, such as those that have resulted in more and more effective cancer screening. This category tends to have its greatest influences on individual-level health outcomes that are highly sensitive to interventions by the health care system. In language consistent with the conceptual framework described in Chapter 6, this set of policies focuses most centrally on downstream targets.
• Public health (broadly defined). This broad category includes a range of programs and policies aimed at improving population health by promoting and supporting healthy behaviors, by eliminating environmental hazards, and by promoting access to preventive interventions. Public health strategies to promote healthy behaviors include communication and education campaigns, strategies that create healthier environments (e.g., building walkable neighborhoods, subsidizing access to healthy foods, limiting portion sizes, restricting advertising of unhealthy products), and laws and regulations (e.g., limits on the density of tobacco outlets, taxation of unhealthy products, restrictions on sales to minors, required use of seatbelts, limits on access to guns). Strategies to eliminate environmental hazards include creating and enforcing air pollution and water quality standards, eliminating lead from housing and the air, and cleaning up hazardous sites. Preventive interventions include facilitating access to screening programs and tobacco cessation programs and implementing harm reduction strategies, such as needle exchange programs. A key feature of public health strategies is that they impact the population as a whole and are focused more on preventing than on treating adverse health outcomes.
• Social and economic policy. This broad category includes policies not specifically directed at health, such as minimum wage laws,

• family leave policies, civil rights legislation, zoning regulations, and tax law. The health-in-all-policies approach recognizes that a broad range of policies outside of the health care and public health sectors—including those focused on civil rights, education, labor markets, housing, income, and transportation—have important health implications. Social and economic policies tend to operate on broad scales (e.g., national, state) and to have their greatest influence on upstream influences on health, including the distal drivers of working-age mortality and disparities therein (House, 2015; Schoeni et al., 2008). Thus, this category of policies best fits with the committee’s conceptualization of macro-level influences on mortality trends and disparities as developed in Chapter 6. Policies targeting upstream social and economic factors may also be especially important for mortality disparities because they tend to focus on vulnerabilities of population subgroups that are due to social and economic inequalities. Research on the macro-level effects on health and mortality of economic and social structure and change is relatively new. Therefore, recommendations for further research in this area are provocative and are intended to spur research that can provide stronger empirical support for social and economic policies to promote the public good of health and well-being.

Finally, it should be noted that these categories of policy influence are often synergistic and mutually reinforcing. For example, public health campaigns focused on efforts to increase rates of adherence to treatment recommendations for chronic disease may work hand-in-hand with health care policy (e.g., broader access to Medicaid) and social and economic policy (e.g., policies that increase employment) to influence reductions in working-age mortality. Similarly, efforts by health care providers to address obesity require changes in the built environment to provide people with access to sidewalks and stores that sell healthy foods, but the ability of people to live in such neighborhoods or to afford healthy foods depends on economic, zoning, and housing policies. Thus the committee uses the above three-category policy and research framework to organize the discussion that follows while recognizing that the assignment of the committee’s recommendations to any single category is imprecise and that some of the recommendations are salient to more than one category.

MEDICAL SCIENCE AND HEALTH CARE ACCESS AND DELIVERY

Background

Advances in medical science have contributed to rising U.S. life expectancy through reductions in mortality rates from key causes of death, especially since 1960 (Cutler, 2004; Cutler, Deaton, and Lleras-Muney, 2006). As discussed in Chapter 9, innovations in drug development and prevention, treatment, and control of chronic diseases explain about one-half of the remarkable decline in cardiovascular mortality seen from 1970 to 2010 (Mensah et al., 2017). At the same time, it should be noted that pharmaceutical innovations and medical treatments can have serious deleterious mortality consequences, best illustrated by the role of the pharmaceutical industry in the opioid epidemic, discussed in Chapter 7. And errors in the delivery of care compromise patient safety and can result in unintended adverse consequences, including deaths (Institute of Medicine [IOM], 2000, 2007a; Sunshine et al., 2019).

Among the greatest successes in recent decades was the introduction of protease inhibitors and highly active anti-retroviral treatment for treatment of HIV infection, which led to pronounced reductions in working-age mortality due to HIV/AIDS starting in the mid-1990s, as documented in Chapter 3. Other medical innovations have greatly improved health and reduced U.S. working-age mortality in recent decades. Examples include advances in cancer screening driven by technological breakthroughs; advances in chemotherapy, radiation therapy, and other oncologic treatment modalities (Armstrong et al., 2016; Smith et al., 2010); improved emergency response systems to stabilize motor vehicle crash victims and quickly transport them to trauma centers, as well as advances in trauma care itself (Nathens et al., 2000); technological advances in the treatment of coronary artery disease and stroke; and new care delivery models to speed transport and emergency treatment of patients with acute coronary events or signs of stroke. Looking to the future, naloxone, an easy-to-administer opioid antagonist, can be used to counter the effects of an opioid overdose and thereby save lives.

Advances in preventive medicine have also occurred. For example, nicotine gum and smoking cessation medications were introduced to help treat nicotine addiction, while the development of vaccines has prevented deaths due to infectious diseases. In many cases, the impact of medical breakthroughs has been amplified by public health initiatives and changes in medical protocols, such as clinical guidelines, checklists, and electronic prompts adopted by health care systems and providers to systematize screening (e.g., for cancer, hypertension, hyperlipidemia, HIV infection);

immunizations; and monitoring of factors implicated in chronic diseases, such as blood pressure, cholesterol levels, and blood sugar.

Medical and pharmaceutical innovations are heavily influenced by public policies. For example, new drug development often derives in part from basic research sponsored by the National Institutes of Health. Drug marketing exclusivity, although temporary, allows pharmaceutical firms to control prices and possibly reap large profits. Yet while many of the drugs developed in this market environment have been beneficial, the high costs of drugs and devices have in some instances led insurers to restrict coverage and patients to forego medications, reducing access to care and dampening the beneficial effects of innovation. Adverse health impacts of drugs have also been observed. Vioxx, for example, a nonsteroidal anti-inflammatory drug developed to treat arthritis, was forced off the market after it led to the death of an estimated 55,000 patients from cardiovascular failure (Biddle, 2007). The opioid epidemic is perhaps the most serious example of the adverse consequences of improper promotion, testing, and approval of prescription drugs. It is essential to understand whether events such as these occurred because of inadequate policies or failure to adhere to existing policies or regulations (e.g., appropriate postmarketing surveillance or communications with health professionals) if such adverse drug impacts are to be avoided.

Health care financing policy has also influenced health outcomes across multiple age groups, including the working-age population, in both positive and negative ways. Health insurers—governmental and private—have been instrumental in promoting the uptake of evidence-based preventive services. Among its provisions, the Affordable Care Act of 2010 requires all health plans to provide first-dollar coverage of all services recommended by the U.S. Preventive Services Task Force. Private health plans have also encouraged mammography screening and other preventive services, and research has clearly shown that waiving out-of-pocket expenses increases the uptake of preventive services (Briss et al., 2000; Newhouse, 1993; Solanki and Schauffler, 1999).

A growing body of evidence suggests that the expansion of Medicaid under the Affordable Care Act has improved health and mortality outcomes for children and adults, including those of working age. Medicaid expansion has been associated with a roughly 9 percent reduction in all-cause mortality among working-age adults exposed to the policy change; this effect appears to be growing with time and is estimated to be saving the lives of thousands of working-age Americans each year (Miller et al., 2019). Another study found that expansion states have experienced a 6 percent reduction in opioid overdose deaths and an 11 percent reduction in heroin-related deaths (Kravitz-Wirtz et al., 2020). Individuals in states that

expanded Medicaid coverage also have experienced better health outcomes relative to those in states that deferred expansion (Antonisse et al., 2018).

In a study comparing Medicaid expansion states (Kentucky and Arkansas) with a nonexpansion state (Texas), expansion was associated with a $337 per capita reduction in annual out-of-pocket spending, significant increases in preventive health visits, and a 23 percent increase in the proportion of respondents who described their health as “excellent” (Sommers et al., 2017). Medicaid expansion has also been associated with improved blood pressure control, but not lower rates of in-hospital mortality among heart failure patients (Cole et al., 2017; Wadhera et al., 2018). Among patients with end-stage renal disease, Medicaid expansion has significantly increased the quality of predialysis care and lowered mortality by 8.5 percent (Swaminathan et al., 2018). Medicaid expansion also appears to be beneficial for surgical patients and has been shown to improve outcomes among patients with diverticulitis, aortic aneurysm, peripheral artery disease, cholecystitis, and appendicitis through earlier hospital admission and more optimal care (Loehrer et al., 2018).

Implications for Policy and Research

The above background overview highlights the important role that medical science and health care, as well as health care access and delivery policies, have played in working-age mortality trends and disparities in recent decades, with largely positive but also some seriously negative consequences. Based on its findings regarding the leading causes of death responsible for increasing working-age mortality, the committee developed policy and research recommendations relevant to providing access to care and treatment for persons at risk of dying from drug poisoning, alcohol-related causes, and suicide; to providing care and treatment to reduce obesity; to addressing other metabolic and cardiovascular conditions; and to instituting regulatory policy to avoid future catastrophes like the opioid epidemic.

As discussed above, there is growing evidence that Medicaid expansion under the Affordable Care Act has led to lower mortality among working-age adults living in expansion states. Accordingly, the committee recommends that those states that have not yet expanded access to Medicaid do so as soon as possible, and that research to analyze the long-term effects of Medicaid expansion on the health and mortality of the working-age population be supported. The committee realizes that Medicaid expansion, like many interventions, requires added public expenditures. Yet existing evidence based on common thresholds for cost-effectiveness shows that program to be cost-effective (see, e.g., Sommers, 2017), suggesting that the added spending is warranted.

RECOMMENDATION 11-1: Given recent findings regarding largely better health and lower mortality among working-age adults who live in states that have expanded Medicaid under the Affordable Care Act, the 12 states that have not yet expanded access to Medicaid should do so as soon as possible. The National Institutes of Health and private foundations should also support research to analyze the long-term effects of Medicaid expansion on the health and mortality of the working-age population.

The committee’s findings also indicate substantial unmet needs with regard to mental health care and substance use treatment. These include provider shortages due to inadequate funding, fragmented delivery systems, the lack of parity in behavioral health plans, and scarce options for the uninsured (Carlo, Barnett, and Frank, 2020). The committee recommends that, as part of addressing the demand side of the U.S. substance use problem, policy makers increase access to and the affordability of quality substance use and mental health treatment (see Chapter 7, Recommendation 7-1), and notes that the expansion of Medicaid (see Recommendation 11-1) would help increase access to such treatment.

Both the National Institute on Drug Abuse (NIDA) and substance use treatment professionals advise that addiction be considered a chronic disease and treated accordingly (Dennis and Scott, 2007; Hser et al., 2015; McLellan et al., 2000; National Institute on Drug Abuse [NIDA], 2005). However, more research is needed on the effectiveness of behavioral health interventions in reducing mental illness and its consequences, on improved methods for delivering mental health and substance use treatment, on harm reduction, and on the extent to which inadequate access to these services has contributed to rising working-age mortality from substance use and suicide (see Chapter 7, Recommendation 7-2). Examples of specific research gaps that require attention include improving behavioral approaches to prevention of drug use relapse, addressing the role of nonsubstance-related conditions in addictive behaviors, and developing better interventions to counter the adverse effects of various social groups in promoting substance use.

Despite increasing medical knowledge on how to improve cardiovascular health (e.g., reduce smoking, improve diet, engage in more exercise) and the development of pharmaceuticals (e.g., statins, antihypertensives) to control chronic cardiovascular conditions, persistent socioeconomic and racial/ethnic disparities in care outcomes persist because many individuals face barriers to care that prevent them from fully benefiting from advances in medical knowledge or receiving recommended preventive services, diagnostic tests, and treatments. The committee therefore contends that efforts to improve implementation systems to overcome these barriers and improve the quality and timeliness of treatments for hypertension, diabetes, and

heart disease may help reduce such disparities and help reverse the stalling of improvements in cardiovascular mortality that began in the 2010s. Translational research and implementation science play an important role in addressing barriers to optimal care faced by individuals and modifying the procedures used by providers, health systems, and insurers to close the gap between recommended care and the care many patients receive. The committee therefore recommends more research to better understand the barriers to effective prevention, diagnosis, and treatment of chronic conditions, with special emphasis on the challenges faced by less-educated and lower-income populations (see Chapter 9, Recommendation 9-4).

PUBLIC HEALTH

Background

History has demonstrated that reliance on access to medical innovations and health care access and delivery aimed at individuals, while important, has a limited impact on reducing or narrowing disparities in mortality. Individual health behaviors, such as tobacco use, account for a large proportion of preventable deaths (Mokdad et al., 2004), but the factors that influence lifestyle behaviors are complex. Decades ago, when the Framingham Heart Study (Levy and Brink, 2005) and other observational studies established the major cardiovascular risk factors—e.g., smoking, poor diet, sedentary behavior, hypertension—the policy response was to urge the public (and counsel patients) to change their behaviors and adopt healthy habits. To identify smokers and counsel them to quit, physicians were advised to systematically complete the “5A’s” (Ask, Advise, Assess, Assist, and Arrange) for every patient (Glynn and Manley, 1995). For most Americans, however, behavior change (and maintenance) proved difficult, especially in combating such powerful habits as smoking and overeating, which were heavily promoted by corporations and advertisers and reinforced by cultural norms. Far too often, those with the determination to change their lifestyle encountered barriers, such as limited access to affordable nutritious foods; outdoor places to exercise, walk, or cycle; counseling programs for smoking cessation or weight loss; and medical care for hypertension. Given the limitations of reliance on individual-level change, public health policy initiatives at the national, state, and local levels have been crucial to support behavior change.

The case of tobacco control policy, perhaps the greatest public health success of the 20th century, is instructive. Specifically, much of the reduction in working-age mortality from tobacco-related diseases (e.g., lung cancer, ischemic heart disease) that is documented in Chapter 3 reflects the nation’s remarkable success in reducing smoking. Fully 42 percent of

American adults smoked cigarettes in 1965, 1 year after U.S. Surgeon General Luther Terry released the landmark report on the dangers of tobacco (Advisory Committee to the Surgeon General of the Public Health Service, 1964; U.S. Department of Health and Human Services [HHS], 2014). The reduction in smoking from 42 percent in 1965 to 13.7 percent in 2018 was achieved in part by a succession of public health policy interventions at the national, state, and local levels, along with other influences, such as social networks that changed the culture around smoking (Pampel, 2002) and early efforts by employers and businesses to discourage smoking. These interventions included not only educational campaigns (e.g., public service announcements, warning labels on tobacco products) but also widespread implementation of laws and regulations aimed at discouraging tobacco use. For example, the suite of policies that reduced smoking rates included cigarette taxes (Hoffman and Tan, 2015), bans on indoor smoking (Frazer et al., 2016), and restrictions on the sale and marketing of tobacco products (particularly to minors) (Harder, 1996).

Exposing the role of the tobacco industry in promoting its products despite knowledge of the addictive properties of nicotine was crucial to continued declines in smoking. Documents disclosed in the 1990s revealed that manufacturers were aware of the health risks posed by tobacco as early as the 1950s (Glantz and Balbach, 2000). The release of these materials fueled legal action by more than 40 states to sue tobacco companies for violating consumer-protection and antitrust laws. The Master Settlement Agreement was reached in 1998, requiring manufacturers to curtail marketing practices and provide perpetual payments to the states for the costs of tobacco-related illnesses (more than $200 billion over the first 25 years) (Myers, 2018). In 2006, in United States vs. Phillip Morris, a federal court held several tobacco companies liable for racketeering; in 2009, a law gave authority to the Food and Drug Administration (FDA) to regulate the tobacco industry.¹ This history would find parallels in recent discoveries about prior knowledge of the addictive properties of prescription opioids in the pharmaceutical industry and new lawsuits by states seeking remedies for losses due to the opioid epidemic.

Although not as successful as tobacco control, policies designed to reduce alcohol use, such as prohibiting alcohol sales before age 21, restricting the density of alcohol outlets, raising the minimum legal drinking age (in the 1970s and 1980s), and adopting “zero tolerance” policies, have been effective (Carpenter, 2004; Carpenter and Dobkin, 2009; Carpenter et al., 2007; O’Malley and Wagenaar, 1991; Sherk et al., 2018; Wagenaar and Toomey, 2002). Actions taken by government and the automobile

___________________

¹ Family Smoking Prevention and Tobacco Control Act of 2009, Pub. L. No. 111-31, 123 Stat. 1776. Available: https://www.congress.gov/111/plaws/publ31/PLAW-111publ31.pdf.

industry—including federal standards for automobile safety, the enactment of laws to enforce speed limits, mandated use of occupant restraints, the prosecution of individuals driving under the influence of alcohol or drugs, and the development of safer vehicles—were central to the reduction in deaths from transport accidents (Byrnes and Gerberich, 2012; Dinh-Zarr et al., 2001; Macpherson and Spinks, 2008; National Academies of Sciences, Engineering, and Medicine [NASEM], 2020b). Even more broadly, public health policies have been important in reducing mortality through a wide range of environmental measures. For example, water fluoridation (Iheozor-Ejiofor et al., 2015) and lead abatement have led to significant health improvements (Wilson et al., 2006). Likewise, the Clean Air Act and associated evidence-based regulation of criteria pollutants resulted in major reductions in contaminants, reducing morbidity and mortality from many diseases linked to air pollution (Samet, Burke, and Goldstein, 2017).

In other areas, however, public health policies and programs have been less effective. For example, 1920s-era alcohol prohibition and the Reagan-era “Just Say No” campaign were ineffective solutions for preventing substance use and addiction (Hornik et al., 2008). “Just Say No” demonized illicit substances and aligned drugs in general with a vaguely defined deviant group, and presented substance use as a collective moral failure of specific communities instead of treating them as a public health issue. The failure of this approach is also illustrated by the ineffectiveness of the Drug Abuse Resistance Education (D.A.R.E) program (Pan and Bai, 2009; Robert Wood Johnson Foundation, 2010), which brings police officers to schools to warn students about the dangers of drugs.

A number of public health policies and interventions have been proposed to create environments that are less obesogenic and are conducive to healthy eating and physical activity. These include conducting educational campaigns (in mass media and at work), implementing menu labeling and providing dietary guidelines, increasing access to healthy foods and limiting access to unhealthy foods (through subsidies and taxes), promoting physical activity through changes in built environments and workplace interventions (focused on walkability, public transport, and standing desks), and passing zoning laws to limit the density of alcohol outlets and fast food restaurants (Community Guide, 2020a; IOM, 2007b, 2012a; Johnston et al., 2014; Lee et al., 2019). These efforts have had success in specific settings but have not always been generalizable, sustainable, or easy to adopt on a national scale (Lee et al., 2019). Yet while many of these efforts have been local and not widespread, policy makers and the food industry share some of the blame for this lack of progress. The U.S. Department of Agriculture, which for decades has promulgated dietary guidelines to promote healthy eating and combat obesity, has also protected the meat and dairy industries (Nestle, 2013) and, since the 1980s, has issued farm subsidies that have generated

an oversupply of corn, thereby lowering its cost and encouraging the production of high-fructose corn syrup, the major sweetener added to the food supply. The widespread use of high-fructose corn syrup and the production of low-cost calorie-dense foods may help explain the marked increase in obesity rates that occurred after these products were introduced (Franck, Grandi, and Eisenberg, 2013).

Policies to address food insecurity and/or the nutrition of vulnerable populations have been effective, helping to reduce health and mortality risks. For example, access in childhood to the Supplemental Nutrition Assistance Program, also known as Food Stamps, reduced the incidence of metabolic syndrome and working-age mortality among persons ages 40–64 (Heflin, Ingram, and Ziliak, 2019; Hoynes, Miller, and Simon, 2015), and the Special Supplemental Nutrition Program for Women, Infants, and Children program achieved success in improving maternal and child health outcomes (Bitler and Currie, 2005; Chorniy, Currie, and Sonchak, 2020; Currie, 2001, 2009).

A general lack of adequate investment in public health and related infrastructure at the national, state, and local levels has been a major challenge in implementing successful public health interventions. A case in point is the nation’s underinvestment in governmental public health agencies, a problem that has been recognized for decades (Bhutta, 2012; DeSalvo et al., 2017; IOM, 2002, 2012b; Mays and Hogg, 2015). This underinvestment has limited the ability of state and local health departments to fully sustain traditional public health services, as well as to create broad and innovative programs to better prevent and control acute and chronic illnesses. This underinvestment has also made the nation more vulnerable to public health crises, hampering the ability to respond adequately to natural disasters, disease outbreaks, or pandemics.

In sum, evidence indicates that the impact of public health initiatives on working-age mortality has been mostly positive, but that these initiatives have not been uniformly successful, and many potential policies (e.g., soda taxes, built environment interventions), although promising, have not been implemented on a broad or national scale. The example of tobacco control is a clear success: the combination of tax policy, smoking restrictions, and public health campaigns clearly altered behavior, improved health, and reduced mortality due to cigarette smoking among working-age adults. Other clear successes include laws targeting air and water pollution and vehicle safety, which in both cases compelled upstream action to improve downstream exposures to health risks. Purely informational campaigns have been less successful but may have laid the foundation of support for stronger policy action in specific cases, such as cancer screening, the prevention and management of HIV/AIDS, and detection and control of hypertension and high cholesterol.

Implications for Policy and Research

The above background overview suggests that there is an important role for public health policies in improving U.S. health and reducing working-age mortality in the future. Based on the findings presented in this report, the committee developed several policy recommendations focused on curbing the availability of addictive drugs and reducing rates of obesity and smoking. The committee also identified several gaps for which additional public health research could help pave the way for future policy recommendations that could help reduce working-age mortality.

Considering the vast loss of life that resulted from the approval, production, distribution, and promotion of opioids and other highly addictive drugs (see Chapter 7), the committee strongly recommends strengthening regulatory control and monitoring of the development and marketing of prescription narcotics. In addition, the committee recommends developing, funding, and enforcing tough internal standards within the pharmaceutical industry, with strong sanctions for violation of these standards. Moreover, the committee recommends that government policy makers at all levels invest in programs focused on substance use as a public health issue and pursue alternatives to arrest and incarceration. These programs should be aimed at reducing barriers to and encouraging entry into substance use disorder treatment (see Chapter 7, Recommendation 7-1).

The United States has already experienced some early successes with such drug-related policies, as discussed in Chapter 7. Guidelines limiting physician prescribing of opioids, monitoring to identify excessive levels of prescribing, and the implementation of “pill mill” laws requiring providers to provide clinical documentation from medical records to support the prescribing of these drugs have been effective in controlling the misuse of prescription opioids (Kiang et al., 2019). Alpert and colleagues (2019) found that states that monitored physicians’ prescribing of opioids and other Schedule II drugs had fewer deaths due to prescription opioids (Alpert et al., 2019). However, there have also been some unintended consequences of the tightening of opioid prescribing, including the substitution of illicit opioids (heroin, fentanyl) for prescription opioids among individuals who had developed addiction or dependence (Hadland and Beletsky, 2018; Mallatt, 2019); rapid and unsafe titration of opioids among chronic pain patients (Kertesz, Gordon, and Satel, 2018); and substantial difficulty among chronic pain patients, including older adults with multiple chronic health conditions, in accessing the opioids needed to manage their pain (Ritchie et al., 2020). Given the preventable loss of life due to the opioid epidemic that unfolded over the past 25 years, the committee strongly endorses more caution with respect to bringing highly addictive and potentially lethal drugs to market. However, policies that reduce access

to prescription opioids without addressing demand are likely to have limited success in reducing drug overdose rates (Ciccarone, 2019; Cicero, Ellis, and Kasper, 2017; Dasgupta, Beletsky, and Ciccarone, 2018; Hadland and Beletsky, 2018).

This report also identifies key knowledge gaps for which more research on supply-side issues is needed. To further support the committee’s policy recommendation for better regulatory control of narcotic prescription drugs, further public health research is needed on the mechanisms that underlie physicians’ and patients’ unintended responses to tighter regulation of drugs posing a high risk of misuse and addiction. Evidence shows, for example, that some individuals who were dependent on prescription opioids were pushed by their inability to obtain those drugs to markets for heroin and fentanyl. Research on strategies for preventing such unintended negative consequences should therefore be conducted in parallel with the development of policies for better regulatory control of narcotic prescription drugs (see Chapter 7, Recommendation 7-3).

The substance use crisis extends beyond drugs. The committee also found considerable evidence of increasing working-age mortality from alcohol-related diseases (see Chapter 4), the causes of which are not entirely clear. The committee therefore urges research to arrive at a better understanding of how changes over time in alcohol consumption, changes in the advertising and promotion of alcohol, and changes in the cultural acceptance of alcohol use have contributed to increases in alcohol-related mortality (see Chapter 7, Recommendation 7-3).

One demand-side argument attributes the trend in mortality due to drugs, alcohol, and suicide to increased vulnerability to substance use brought on by both long-term underlying and recently increasing stress, distress, “despair” (for which there are currently no clinically validated measures), pain, and mental illness (particularly depression). However, fundamental data are lacking on the epidemiology of mental health to confirm whether stress, distress, despair, pain, and depression have definitively increased, let alone to link those increases to trends in drug or alcohol use. A substantial literature documents that persons with mental illness are at increased risk of substance use disorders (NIDA, 2020; Unger et al., 1997; Volkow, 2001). Although mental illnesses and substance use disorders are therefore closely intertwined, it has been difficult for both researchers and policy makers to understand trends and disparities in stress, distress, despair, pain, and depression and how such trends may be related to substance use and substance use disorders. The committee therefore recommends further research on physical pain and the various psychosocial indicators that increase and/or decrease the risk of unhealthy behaviors related to substance use.

RECOMMENDATION 11-2: Federal agencies, in partnership with private foundations and other funding entities, should support research that tracks physical pain and the various psychosocial indicators, including stress, distress, despair, hopelessness, coping, resilience, and grit, that increase and/or decrease the risk of unhealthy behaviors related to substance use at the population level; explores relationships between these indicators and various causes of mortality and morbidity; and examines how trends in these indicators and their associations with mortality and morbidity vary by demographic group, socioeconomic status, and geography.

The rise in suicide mortality among Whites, especially White men, documented in this report occurred in the context of changing means of suicide. Despite evidence that more firearm-related suicides occur in states with looser gun regulations and greater gun ownership, the proportion of all suicide deaths related to firearms declined from 1990 to 2017, while the proportion due to hanging, suffocation, and strangulation increased. Thus, more needs to be known about means of suicide to better understand the increase in different modalities, how they differ by sex, and what factors might precipitate the choices made in this regard. Research on the role of gun control laws and gun availability in suicide mortality is particularly warranted, with attention paid to the causal effect of changes in gun control laws and gun availability on trends in suicide mortality (see Chapter 8, Recommendation 8-1). Given the different levels, trends, and disparities in mortality due to drug poisoning, alcohol, and suicide, the committee examined each of these causes of death separately rather than grouping them into a single category of “deaths of despair.” Nonetheless, research is needed to test more effectively whether there are important drivers of mortality that may be common across these three causes. Specifically, the committee recommends public health research to explore how the various mechanisms that explain sociodemographic and geographic differences and temporal changes in mortality due to drug poisoning compare with those that explain sociodemographic and geographic differences and temporal changes in mortality due to alcohol and suicide (see Chapter 7, Recommendation 7-3).

Chapter 9 provides important explanations for the troubling stall in declines in mortality (and recent mortality increases among some demographic subgroups) due to cardiometabolic diseases among working-age adults. The key explanation for these trends is the increasing prevalence of adult obesity, coupled with long-term exposure to obesity that all too often begins in childhood, especially among recent cohorts (Abdullah et al., 2011; Owen et al., 2009; Stokes, Ni, and Preston, 2017). Chapter 9 also highlights disparities in obesity across population subgroups—including by race and ethnicity, gender, and socioeconomic status—that influence

cardiometabolic disparities in mortality (Hales et al., 2020; Lee, Harris, and Gordon-Larsen, 2009). Public health policy can play an important role in reducing obesity, with particular attention to curbing trajectories of potentially problematic weight gain in childhood and adolescence. Accordingly, the committee recommends that obesity prevention programs start early in life and be targeted to children and adolescents most at risk (e.g., racial/ethnic minority groups, females, and people living in poverty and neighborhoods of low socioeconomic status) and those who are overweight or gaining weight to intervene before obesity trajectories become set throughout the life course (see Chapter 9, Recommendation 9-3).

As discussed in Chapter 9, however, the causes of obesity are multifaceted and therefore difficult to address with single public policy initiatives, a point underscored by the limited headway made in reducing obesity over the past three decades (IOM, 2012a; Ravussin and Ryan, 2018). More research is needed to identify the multilevel and interactive causes of obesity to support the development of a multipronged public policy approach for addressing this major public health problem with as much success as was achieved through tobacco control policies and public health programs. Almost all obesity scholars point to the important role of obesogenic factors in the physical and food environments, including the interplay between individual health behaviors involving diet and physical exercise and societal-level changes in food production, transportation systems, green space, and sedentary work environments.

For example, there is evidence that technological changes in the way food is produced, distributed, and consumed have contributed to the increase in obesity, and that there is a role for public health policy in improving the production of healthy foods and reducing the distribution and consumption of unhealthy foods, especially among children and adolescents. The committee’s findings also document the success of healthy diets in achieving weight loss, at least for the short term, and the fact that regular exercise is almost always beneficial in reducing overweight and obesity. However, efforts to promote lifestyle changes involving healthy diets and regular exercise (e.g., through worksite health promotion programs and free gym memberships) may not always be effective or sustainable. Thus, research is needed to explore the factors that erode short-term success in diet and exercise changes and, conversely, the factors that promote long-term lifestyle change to reduce obesity. A focus of this research should be on environmental drivers (e.g., occupational activity, exposure to chemicals, food deserts, green space and walkability, economic inequality, residential segregation, duration of exposure to electronic screens, advertising, the increase in “hyperpalatable” processed foods). Specifically, the committee recommends research evaluating the effectiveness of programs and policies that promote the consumption of healthy foods and adoption of healthy

lifestyles, as well as those that discourage the consumption, manufacturing, and advertising of highly processed and poor-quality foods and unhealthy lifestyles. The committee also recommends considering how systemic changes in food production, workplace systems, urban design, and transportation and other societal-level changes have fostered and sustained obesogenic environments and sedentary lifestyles to determine how those environments have deleterious consequences for population health (see Chapter 9, Recommendation 9-1).

As noted above, research on environmental impacts is complex, and the wide range of study designs, methods, and environmental variables involved makes it difficult to identify causal pathways. As a first step toward identifying some of the key drivers, the committee recommends research using experimental designs and taking advantage of existing neighborhood experimental projects (e.g., Moving to Opportunity) to examine the causal role of obesogenic factors in the environment and determine which are most responsible for the rise in obesity prevalence and body mass index levels (see Chapter 9, Recommendation 9-2).

The food industry—aided by legislation and budget decisions promoted by lobbyists and politicians from agricultural states—has also contributed to the obesity epidemic by successfully leveraging advertising and marketing techniques to boost consumption of calorie-dense foods (Charlebois, Tamilia, and Labrecque, 2007; Freudenberg, 2014; Nestle, 2013). As noted earlier, the farm subsidies authorized by Congress in the 1980s encouraged the oversupply of corn, thereby lowering its cost and encouraging the production of high-fructose corn syrup, a major sweetener that has become ubiquitous in American processed food. U.S. per capita consumption of high-fructose corn syrup increased from 0.8 g per day in 1970 to 91.6 g in 2000 (Bray, Nielsen, and Popkin, 2004). In addition, restaurants promoting inexpensive, unhealthy, and “all you can eat” menus have proliferated, especially in socioeconomically disadvantaged neighborhoods most at risk for obesity.

The need for solutions is widely recognized in the public and private sectors, driven not only by public health concerns but also by the threat obesity poses for employers, the business community, and the armed services. Further work is needed to build on recent efforts—some led by the food industry itself and others by public health authorities—to discourage the production and purchase of unhealthy foods or at least give consumers better information with which to make healthier food choices. Examples include self-regulation of or restrictions on misleading advertising (Graff, Kunkel, and Mermin, 2012), pricing and tax strategies (Blecher, 2015), “Nutrition Facts” product labels mandated by the FDA (U.S. Food and Drug Administration [FDA], 2020), menu labeling by restaurants (VanEpps et al., 2016), zoning decisions to reduce the proliferation of fast food restaurants, and other measures. Evaluating policies and initiatives that affect the diets

of young people is crucial, including replacing corn syrup with natural sweeteners, taxing sugary drinks and soda, restricting advertising targeting children, and removing vending machines from schools. That U.S. caloric intake per capita outpaces that of other high-income countries (Institute of Medicine and National Research Council [IOM and NRC], 2013) suggests the need to identify and address structural causes for that U.S. distinction, from more permissive regulation of the food industry and advertisers to cultural differences in lifestyle (see Chapter 9, Policy Conclusion 9-1).

Although tobacco control policy has been a great U.S. public health success, reductions in smoking behavior have been much greater among the highly educated than among the less educated, leaving large disparities in tobacco use that in turn contribute to disparate rates of mortality from cardiometabolic diseases and cancer. The committee recommends further public health research to address the barriers to smoking cessation and prevention of initiation faced by populations that continue to smoke at high rates, especially those with less education or income, and to evaluate programs that have been successful in promoting smoking cessation or preventing initiation (see Chapter 9, Recommendation 9-4).

In understanding and addressing public health problems and diseases, it is also important to take account of important comorbidities and co-occurring conditions in the absence of which a person may not have died (e.g., alcohol or drug involvement in motor vehicle or pedestrian accidents, chronic substance use and heart disease, injection drug use and bloodborne infectious disease). The COVID-19 pandemic has vividly illustrated the critical role of comorbidities: those who have obesity, autoimmune diseases, hypertension, or heart disease have died at much higher rates from COVID-19 infection relative to those without such conditions. Most approaches to coding and describing trends in mortality (including the approach used in this report) assign a single cause of death—the underlying cause, defined as “the disease or injury which initiated the train of morbid events leading directly to death, or the circumstances of the accident or violence which produced the fatal injury” (World Health Organization [WHO], 2011, p. 31). However, most deaths involve numerous contributing, sequential, and overlapping conditions. Assigning a single cause of death is therefore a simplistic and somewhat artificial way to describe and understand mortality trends and their explanations.

RECOMMENDATION 11-3: Researchers should more frequently use the International Classification of Diseases (ICD)-10 codes for multiple causes of death in their examination, analysis, and explanations of mortality trends and disparities in order to better identify the various factors that act together in causing death and how those factors and their combinations change over time.

SOCIAL AND ECONOMIC POLICIES

Background

Many medical and public health policies discussed above focus on improving health one patient or one disease at a time. In contrast, social and economic policies typically are not designed to reduce mortality for any or all causes of death but may affect survival nonetheless. Policies outside the sectors of medicine and public health that influence education, jobs, income, wealth, housing, racial/ethnic inequalities, immigration, and other determinants of health may also influence trends and disparities in working-age mortality. Macro-level social and economic policies—enacted by cities, states, and the federal government—have the potential to alter the distribution of key societal resources (e.g., income, education, housing, wealth). They also can reduce or exacerbate social inequities (e.g., by gender, race and ethnicity, immigrant status, sexual minority status, socioeconomic status, and ability status) and disparities in health and mortality (House, 2015; Hummer and Hamilton, 2019; Phelan, Link, and Tehranifar, 2010; Schoeni et al., 2008).

Researchers and policy makers face key challenges when considering social and economic policies (see Chapter 10 for discussion of empirical work on the relationship between economic factors and mortality). First, the effects of such policies on mortality are difficult to measure because the policy initiatives tend to be quite distant from the more proximate causes of mortality (e.g., stress, obesity, smoking) and operate mainly through indirect pathways (e.g., changes in social and economic conditions). Second, changes in social and economic conditions may not be due exclusively to policy changes. For example, the downsizing of the U.S. manufacturing sector that claimed so many jobs was fueled in part by factors largely outside the control of policy makers, such as the growth in automation and information technologies and other global market forces. The significant potential for confounding further makes it difficult to infer causality from associational studies of policy contexts and mortality trends and disparities. For example, trends in death rates due to drug poisoning differed between the United States and European countries over the past 25 years even though recessions and economic disruptions due to globalization that were experienced by the United States also affected European countries. This divergence may well be due to U.S.–European policy differences in such areas as social welfare programs and universal health care that prolonged the economic pain for families in the United States. However, the evidence is not rich enough to disentangle the extent to which the more favorable experience in Europe reflects differences in policies related to economic

security and unemployment, differences in medical culture related to use of pain medications, or some other factors unrelated to policy.

With the challenges of social and economic policy in mind, the discussion below first describes a few examples of policies related to broader social and economic well-being that have been demonstrated to influence mortality in the United States, and then summarizes policy and research implications of social and economic factors.

As a first example, studies have shown that some income-related policies influence population health and working-age mortality. Research has demonstrated that the Earned Income Tax Credit (EITC) increases employment and income and improves a variety of health outcomes (Hamad and Rehkopf, 2015; Hotz, 2003; Hoynes and Rothstein, 2016; Hoynes, Miller, and Simon, 2015; Komro et al., 2016; Strully, Rehkopf, and Xuan, 2010).² Evidence is also suggestive that minimum wage laws improve economic security and health outcomes³ and may have impacts on working-age mortality (Andreyeva and Ukert, 2018; Dow et al., 2019; Kaufman et al., 2020; McCarrier et al., 2011; Narain and Zimmerman, 2019; Rosenquist et al., 2020; Tsao et al., 2016; Van Dyke et al., 2018). More broadly, policy efforts aimed at increasing educational attainment among the U.S. population have been shown to have effects on reducing mortality rates decades later (Halpern-Manners et al., 2020; Lleras-Muney, 2005). Other social and economic policies focused on unemployment benefits, the provision of food, preschool education and childcare, parental leave, housing, community development, economic growth and inflation, and taxation hold great potential for influencing health and reducing working-age mortality. The many policy options that fit into these domains, however, are rarely envisioned as health policies, and little empirical work has directly assessed their effects on trends and disparities in working-age mortality (House, 2015).

Many social and economic policies, especially those targeting populations of lower socioeconomic status, have been helpful in addressing racial/ethnic inequities. As noted earlier, as a result of the U.S. history of structural racism and related discriminatory practices that persist today, non-Whites—and particularly Blacks and American Indians—have faced greater barriers to education, good jobs, high incomes, and stable housing relative to Whites and have had fewer opportunities than Whites to transfer wealth to subsequent generations. Policies aimed at reducing such inequities, which played a historic role in advancing civil rights, have been

___________________

² These outcomes include children’s cognitive abilities, birthweight, Apgar score, child development, fertility, cognitive development, mortality, natality, and postneonatal mortality.

³ These outcomes include postneonatal mortality, heart disease death rate, suicide death rate, self-reported hypertension, self-reported fair or poor health, and self-reported mental and physical health.

associated with improved health and mortality outcomes among Blacks. For example, research has demonstrated an association between the Head Start program and other civil rights legislation of the 1960s and subsequent reductions in working-age mortality among Blacks (Almond and Chay, 2006; Johnson, 2019; Kaplan, Ranjit, and Burgard, 2008). Chetty and colleagues (2016) found that policies to improve housing opportunities in segregated communities, such as the Moving to Opportunity program, were associated with improved social and health outcomes for low-income and minority individuals. Unfortunately, however, Blacks have continued to experience disproportionately high mortality rates, and, as discussed in Chapter 3, progress in lowering working-age mortality and narrowing the Black–White mortality gap has now stalled. Indeed, during the 2010s, working-age mortality increased among all non-White populations, reversing decades of progress (Woolf et al., 2018).

Findings presented in this report provide ample evidence that the recent rise in U.S. working-age mortality was concentrated among socioeconomically disadvantaged populations, particularly those with a high school degree or less, and that U.S. working-age mortality remains substantially higher among Blacks and American Indians, in particular, compared with Whites. There is also reasonable evidence that some targeted social and economic policies, such as higher minimum wages and the EITC, can reduce mortality rates in these vulnerable populations (Dow et al., 2019; Kaufman et al., 2020). Evidence on investments in education-, housing-, and employment-related policies have also been shown to reduce racial/ethnic discrimination and racial/ethnic disparities in health and mortality (Williams, Lawrence, and Davis, 2019). Thus, while the associations between social and economic policy and working-age health/mortality remain ripe for research, the committee’s findings also carry important policy implications.

Implications for Policy and Research

The committee’s findings demonstrate that the largest increases in working-age mortality since 1990 occurred among Whites. At the same time, however, these findings reveal that 2010 was an inflection point when all-cause mortality rates also began to increase among other racial/ethnic groups of working-age adults. Thus, recent trends in working-age mortality are problematic for all racial/ethnic groups in the United States. The committee furthermore identified large and in some cases widening socioeconomic and geographic disparities in working-age mortality. The increases in mortality rates that occurred over the past 25 years among persons with a high school degree or less and in many nonmetropolitan portions of the country are urgent concerns that require greater focus by the research community. Many of these racial/ethnic, socioeconomic, and

geographic disparities in mortality are likely to be strongly influenced by macro-level (e.g., social, economic, cultural, policy) factors, both historical and contemporary.

While the explanations for high and rising working-age mortality discussed in this report are often specific to certain causes of death, they reveal some common underlying themes that affected particular population groups at different time periods or in different contexts.

The first of these themes relates to adverse economic trends (e.g., stagnant wages, industrial shifts, job losses) that affected certain geographic areas and population groups more than others. The loss of manufacturing, mining, and other jobs in the industrial Midwest and Appalachia in the 1970s–2000s led to long-term economic decline and associated social problems in these areas, particularly among White individuals without a 4-year college degree. Scholars have noted that Blacks in urban centers experienced similar economic and social transformations in earlier decades when manufacturing jobs left the cities; unemployment rates increased, particularly among males; and marriage rates plummeted (Cherlin, 2009, 2018; Torr, 2011; Wilson, 1987). The decline in economic conditions among Blacks that preceded the increase in drug-related mortality during the cocaine and heroin epidemics of the 1970s and 1980s is not thematically unlike the loss of jobs in suburban and rural areas among Whites that preceded the opioid epidemic in the 1990s and 2000s. The different governmental and policy responses to these epidemics are telling, however, revealing a deeper societal explanation not only for these discrepant policies but also for the persistent Black–White mortality gap: the drug crisis among Blacks of low socioeconomic status was treated primarily as a criminal justice problem, while the crisis among their White counterparts was largely recognized as a public health problem, although the criminalization of addiction remains strong today.

A second theme—which helps explain the pace and timing of rising 21st-century working-age mortality and long-standing racial/ethnic disparities in mortality that have persisted throughout U.S. history—is socioeconomic inequality. Inequality is defined as a state of unequal access to opportunity, resources, or means in a process leading to health and longevity (Braveman and Tarimo, 2002; Krieger, Williams, and Moss, 1997; McCartney, Collins, and Mackenzie, 2013). As outlined in the conceptual framework presented in Chapter 6, the magnitude and forms of socioeconomic inequalities experienced by social groups vary based on certain characteristics (e.g., race and ethnicity, age, sex, gender identity, sexual orientation) and operate at multiple levels (e.g., institutional, community, family, individual). Moreover, socioeconomic inequality is structured across groups according to entrenched social hierarchies (Massey, 2007). As a result, groups that have systematically been treated unfairly in society

because of racism or other forms of discrimination possess far fewer socioeconomic resources relative to advantaged groups (e.g., the White majority), creating large socioeconomic inequalities.

Socially patterned disparities in health and mortality reflect the downstream outcomes of these unequal upstream processes (Link and Phelan, 1995; Phelan, Link, and Tehranifar, 2010; Williams and Sternthal, 2010). Understanding the causes of inequality can enable policy makers to derive solutions by intervening on those causes to facilitate the goal of health equity, both within and among groups. As a consequence of the long history of structural racism in the United States (more on this below), Blacks and American Indians, in particular, have experienced long-standing and persistent inequalities in opportunities for educational attainment in high-quality schools, stable jobs with good incomes, wealth accumulation, and the kind of intergenerational mobility that would place them on socioeconomic parity with Whites. In recent decades, socioeconomic inequality has also deepened among Whites and within U.S. society as a whole.

The growing importance of education within U.S. society and of academic credentials in obtaining well-paying technical and professional jobs has left those without a college degree with fewer opportunities for stable employment, employment-related health benefits, and social mobility. After World War II, White males without a college degree often had opportunities for jobs in blue collar industries, with middle-class salaries, health benefits, and pension plans; to a lesser but still important extent, this was also true for Black males. However, the decades that followed the 1970s and 1980s and the shift from a manufacturing to a service economy brought growing inequality between less- and more-educated American workers, a trend that in some ways redefined the position of low-educated Whites in the U.S. social stratification system, with profound consequences for their health and longevity. For example, evidence presented in this report indicates that mortality from substance use increased more rapidly among less-educated relative to more highly educated White adults throughout 1990–2017 (Case and Deaton, 2015; Denney et al., 2009; Geronimus et al., 2019). Accompanying these trends is evidence of the social toll of rising inequality among less-educated Whites, who experienced greater social isolation, family breakdowns, and declining institutional support systems (Case and Deaton, 2020; Cherlin, 2014, 2018; Ruggles, 2015; Torr, 2011; Wilcox et al., 2012).

A third theme that emerges from the committee’s explanations for the trends and disparities in working-age mortality is vulnerability, which mediates the degree to which adverse economic conditions and socioeconomic inequality make some groups more susceptible than others to morbidity and mortality risks. For example, as a result of educational, job, and housing discrimination, Blacks tend to work and live in segregated disadvantaged

neighborhoods, which increases their exposure to obesogenic, unsafe, and low-resource environments that lack mental and physical health care and increase their mortality risks. Today’s drug overdose crisis emerged from a “perfect storm” in which the flooding of the market with highly addictive yet deadly products occurred as the population was growing more vulnerable to physical and emotional pain, generating a heightened demand (and market) for substances that could temporarily bring relief. This demand was initially met by increased availability of prescription opioids, prescribed predominantly to Whites. Restrictions on opioid prescribing, however, increased the demand for and supply of illegal drugs, which became more accessible to other vulnerable groups—working-age Blacks, Hispanics, and low-educated Whites who lacked access to prescription opioids. Indeed, rates of overdose involving heroin, fentanyl, and cocaine started to rise for all racial/ethnic groups after 2010.

Declining economic conditions, socioeconomic inequality, and vulnerability are themes that help in understanding how the different and changing social, economic, and geographic contexts of population subgroups may explain recent trends and disparities in working-age mortality. Macro-level shifts in economic conditions and inequalities that operate at all levels of society have made various subgroups vulnerable in different places and times. For example, much attention with respect to the widening educational disparities in mortality among Whites has focused on the shifting fortunes of less-educated Whites that have upended their social and economic position. Blacks, American Indians, and people of Hispanic and Asian descent have endured such vulnerabilities as being victims of enslavement, genocide, and discrimination for centuries. The “American Dream” that children would live a better life than their parents has often been out of reach for Blacks and American Indians, but in recent decades has also eroded for working-class Whites. Understanding the pathways by which racial/ethnic, socioeconomic, and geographic disparities in mortality occur and perpetuate is an area in need of additional research focusing on the upstream macro-level historical and contemporary drivers and the downstream processes through which they operate to create health inequities.

RECOMMENDATION 11-4: The National Institutes of Health and other government and private research funders invested in understanding the structural and policy determinants of health should support a robust research program aimed at identifying the macro-level historical and contemporary drivers (e.g., social, economic, cultural, policy) of health and mortality inequities and the mediators (e.g., environmental, socioeconomic, health care, biological, psychological, behavioral) through which these drivers operate to create and sustain persistent racial/ethnic, socioeconomic (income and education), and geographic

(including rural–urban, regional, and across- and within-state) disparities in U.S. working-age mortality. Particular emphasis should be on understanding policy solutions that may be effective in reducing and eliminating inequities in health and well-being.

While, as noted above, Whites have experienced increasing socioeconomic inequality in recent decades, very wide socioeconomic inequalities remain among racial/ethnic groups that favor Whites and have always disadvantaged minority groups. Blacks and American Indians/Alaska Natives (AI/ANs), in particular, have experienced institutional- and individual-level injustices for centuries, brought about by the long history of structural racism in U.S. society that continues to the present day (Alexander, 2010; Du Bois, 1899; Massey and Denton, 1993). The legacy and persistence of structural racism are reflected in the findings presented in this report, which show that working-age mortality rates were much higher among Blacks than among Whites throughout the 1990–2017 period. Although the quality of mortality data for AI/ANs and Asian Americans is problematic, summary findings suggest that AI/ANs have the highest mortality rates of all working-age adults, while Asian Americans have the lowest (see Annex Figure 4-1 and the table in Box 4-2 in Chapter 4). While the Black–White disparity in all-cause mortality narrowed between 1990 and 2017 as mortality rates generally decreased among working-age Blacks and stagnated or increased among working-age Whites, Black all-cause mortality rates across the working ages remained 33–54 percent higher than the rates for Whites in 2017. The Black–White disparity in mortality is so great that 400,000 additional Whites would need to have died each year for the rates for the two groups to be equivalent (Wrigley-Field, 2020). Reflecting this disparity, life expectancy at birth for Blacks was 3.9 years shorter than that for Whites in 2018 (Murphy et al., 2021).

Racial/ethnic disparities in working-age mortality differ across causes of death. While drug mortality rates increased for White males and females of all working ages during 1990–2017, with especially pronounced increases after 2010, Black males ages 55–64 had higher drug mortality rates than Whites throughout the study period. The steepest increases in suicide mortality occurred among Whites of all working ages, especially White males (although Black and Hispanic suicide rates began to increase after 2010). Among cardiometabolic causes of death, Whites of all working ages also experienced increasing mortality due to endocrine, nutritional, and metabolic (ENM) diseases and hypertensive heart disease from 1990 to 2017, while Black mortality from these diseases fluctuated and remained relatively flat over this period, the exception being younger Black adults (ages 25–44), for whom the rates increased after 2010. Even with these fluctuations, however, the Black mortality rates for ENM diseases and hypertensive heart

disease were two to five times higher than the White rates across all working ages and time periods. Blacks relative to Whites experienced a faster pace of decline in mortality from ischemic heart disease and other circulatory diseases from 1990 to 2010, but when mortality improvement stalled for all racial/ethnic groups, Black mortality rates remained almost twice as high as those of Whites in 2017. Because cardiometabolic mortality includes the leading causes of U.S. deaths today (e.g., heart disease, diabetes, stroke), it contributes the largest number and largest proportion of all deaths to U.S. mortality trends. Thus, the persistent Black–White disparity in cardiometabolic mortality is an important driver of the long-standing racial/ethnic gap in all-cause mortality that has disadvantaged Blacks throughout U.S. history. While data limitations prevented an in-depth examination of cause-specific trends in AI/AN working-age mortality, previous research suggests similarly heightened mortality due to cardiometabolic diseases for this group as well.

Structural racism remains a central explanation for the persistent disparity in mortality rates between Blacks and Whites. Structural racism can be defined as the systematic restriction of societal resources (e.g., high-quality schooling, stable work and fair pay, safe housing and neighborhoods, wealth, prestige, respect, freedom) through processes of exploitation, exclusion, normalization, and legitimization that routinely advantage Whites while producing cumulative and chronic adverse outcomes for people of color (Gee and Ford, 2011; Hummer and Hamilton, 2019; Phelan and Link, 2015). As an upstream macro-level factor, the impact of structural racism on health and mortality is challenging to study because of its systemic nature, which manifests in multiple societal institutions, policies, and environments, making it difficult to measure and statistically quantify. As noted earlier, moreover, structural racism operates on health and mortality through a multitude of downstream community and individual-level mechanisms that interact in and across time and space (Gee and Ford, 2011; Hummer, 1996; Phelan and Link, 2015; Williams and Collins, 2001; Williams and Mohammed, 2013) (see Figure 6-1 in Chapter 6).

Structural racism produces racial/ethnic inequalities through two primary mechanisms. The first and largest is limiting access to resources for achieving socioeconomic status, measured by educational attainment, occupational prestige, income, home ownership, and wealth (Gee and Ford, 2011; Hummer and Chinn, 2011; Phelan and Link, 2015). An illustrative pathway through which structural racism affects cardiometabolic mortality is residential segregation (Williams and Collins, 2001; Williams and Mohammed, 2013). Substantial evidence documents that patterns of discrimination in education, employment, income, access to credit, and the real estate industry relegate Blacks to residential environments that provide less access to healthy foods, space for exercise and leisure activities, clean

air and water, and safe and efficient transportation systems—all of which are important risk factors for obesity. Thus as discussed earlier, the legacy and persistence of structural racism contribute to the higher prevalence of obesity among Blacks and their higher rates of mortality from ENM diseases and hypertensive heart disease compared with Whites, whose death rates from these causes have been rising slowly but are still much lower than those of Blacks.

The second mechanism by which racism affects health is biological processes that damage multiple body systems through exposure to overt discrimination, violence, and the daily microaggressions experienced by people of color. The chronic stress of and ongoing vigilance for such experiences harm the neuroendocrine and immune systems and promote unhealthy coping behaviors (e.g., smoking, substance use) (Geronimus et al., 2010; Hicken, Lee, and Hing, 2018; Jackson, Knight, and Rafferty, 2010; McEwen and Lasley, 2002; Williams, Lawrence, and Davis, 2019). These harms can be experienced by people of color at all levels of education or social standing, which helps explain why racial/ethnic disparities in working-age mortality persist even after adjustment for socioeconomic status (Geruso, 2012; Hummer and Chinn, 2011; Williams, 1999; Williams, Priest, and Anderson, 2016).

The structural racism that produces these adverse health effects has a long history in the United States and is considered systemic because that history has left a deep imprint on so many of the policies and practices of the nation’s institutions. Policies enacted decades ago (e.g., redlining, the practice originating in the 1930s of restricting home loans in Black communities) continue to have persistent effects, such as limiting access to resources for social and economic mobility among people of color (Rothstein, 2017). Such policies are not just a vestige of the past; the disadvantages produced by historical policies are compounded by modern policies that continue to be adopted by institutions at the national, state, local, and neighborhood levels in ways that systematically disadvantage people of color.

The committee recognizes that structural racism is too complex to be amenable to easy fixes; isolated interventions are therefore unlikely to achieve meaningful impact. As in other countries, such as South Africa decades ago, success in dismantling structural racism in the United States will invariably require a suite of solutions that confront the problem at multiple levels of society and across sectors. Racial/ethnic disparities in health outcomes will likely persist until this occurs. Given the magnitude of the racial/ethnic mortality disparities documented in this and other reports and the multidimensional institutional- and individual-level factors that help explain them, the committee realizes that, as with structural racism itself, one or even a small number of specific policy levers will be insufficient to eliminate racial/ethnic disparities in working-age mortality. Thus,

the committee’s policy conclusion in this area is intentionally broad and is directed at macro-level (local, state, federal) factors, reflecting the long-term, institutionalized, and multifaceted racism and associated inequalities that perpetuate racial/ethnic disparities in working-age mortality.

POLICY CONCLUSION 11-1: To reduce and ultimately eliminate racial/ethnic and other socioeconomic inequalities that continue to drive racial/ethnic disparities in U.S. working-age mortality, policy makers and decision makers at all levels of society will need to dismantle structural racism and discriminatory policies of exclusion (in such areas as education, employment and pay, housing, lending, civic participation, criminal justice, and health care) and be intentional in ensuring that new social and economic policies serve to eliminate, and not perpetuate, the social and economic inequalities to which racial/ethnic minority groups have long been exposed.

Efforts to revitalize the communities that have been hit hardest by the substance use and overdose crisis will similarly need to address the larger economic and social strains and dislocations that made these communities vulnerable to opioids and other drugs over the past four decades. Doing so may require a holistic approach to community development that involves federal, state, and local governments as well as a range of private-sector actors (see Chapter 7, Policy Conclusion 7-1).

The explanatory chapters of this report on drug- and alcohol-related, suicide, and cardiometabolic mortality all, in some ways, implicate stress as a possible important mechanism contributing to increased mortality from these causes of death. Decades of research have identified chronic stress as a key determinant of health and racial disparities therein for such cardiometabolic conditions as hypertension (Williams, 1999), elevated body mass index and obesity (Hicken, Lee, and Hing, 2018), diabetes (Jackson, Knight and Rafferty, 2010), and heart disease (Williams and Jackson, 2005; Williams, Lawrence, and Davis, 2019). Chronic stresses may also play an important role in helping to explain trends in working-age mortality due to suicide and drug and alcohol use. Social, economic, technological, and cultural shifts over the past 50 years have profoundly changed the family, work, and community environments underlying daily life in the United States, potentially leading to increased stress. Chronic daily stress may be felt more intensely among certain subgroups of the population, especially those with fewer available resources for dealing with such stress in healthy ways. For example, Chapter 3 presents evidence that the increase in working-age mortality during the study period (1990–2017) was more pronounced and geographically widespread among U.S. women relative to men. Women in the United States have shorter lifespans and poorer health

outcomes compared with women in other high-income countries (IOM and NRC, 2013), a disadvantage that some have attributed to the unique stresses that U.S. life imposes on women (Montez et al., 2015). A research program focused on the stressful lives of Americans—one that includes uncovering the sources and the consequences of that stress—would transcend individual causes of death and could provide key insights into both high overall working-age mortality and disparities therein.

RECOMMENDATION 11-5: Given the potential connection of daily stressors to substance use, suicide, and cardiometabolic disease and mortality, federal agencies, in partnership with private foundations and other funding entities, should support research that documents the sources of increasing stress in the lives of Americans (e.g., student debt, foreclosures, job instability, economic insecurity, family instability) and identifies those groups most affected by increasing stress (e.g., the poor, immigrants, young adults, racial/ethnic minorities, women, those living in rural areas, the long-term unemployed, those without a 4-year college degree).

More generally, much more research is needed on the population health and mortality crisis that has been affecting working-age Americans for at least the past 25 years. Widespread recognition of the increase in working-age mortality occurred as recently as late 2015 (Case and Deaton, 2015). While much has been learned since then, as documented in this report, the problem is of such magnitude that two additional streams of research—socioecological and cross-national—are urgently needed to focus squarely on this crisis.

RECOMMENDATION 11-6: Federal agencies, in partnership with private foundations and other funding entities, should support quantitative and qualitative interdisciplinary research on how factors defined at multiple levels (e.g., nation, state, community, family, individual) relate to working-age mortality, especially to deaths involving drug and alcohol use, suicide, and cardiometabolic disease.

• This research should examine how mortality due to drugs and alcohol, suicide, and cardiometabolic disease varies by individual-level demographic characteristics (including sex, race and ethnicity, and socioeconomic status), economic and social factors (e.g., social integration, unemployment, income inequality, public policy), and various levels of geographic characteristics that may change over time (e.g., geographic characteristics of counties, state and local jurisdictions, labor markets, and neighborhood environments).

• The research should be designed to uncover protective and predisposing factors unique to specific population subgroups that can inform policies designed to reduce disparities in working-age mortality.
• The research should explore how mortality is affected by long-term changes in the economy (e.g., changes in employment, employment opportunities, and job characteristics), especially in certain geographic areas; by interaction between economic factors and such social factors as family structure, community support, and religiosity; by the duration of economic hardship; and by programs designed to alleviate economic deprivation and other social stressors.
• The research should consider study designs, measurement strategies, and analytic methods that can strengthen causal inferences and conclusions. Examples include well-designed longitudinal cohort studies with individual-level data linked to time-varying environmental data measured at multiple levels (e.g., states, neighborhoods, families), and approaches that capitalize on natural or quasi-experiments that can be leveraged to identify etiologic (causal) factors and policy impacts.

Finally, some broad work in the area of U.S. health and mortality contrasts social and economic policy regimes in the United States with those in European and other high-income countries around the world. Much of this work suggests that differences in population health and mortality between the United States and its high-income peers may be due to weaker social and economic policy supports for individuals in the United States (Avendano and Kawachi, 2014; Bambra and Beckfield, 2012; Beckfield and Bambra, 2016; House, 2015; Hummer and Hamilton, 2019; IOM and NRC, 2013). To date, however, this promising area of inquiry has not been explored as rigorously as should be the case. In addition, this area of research should include state- and substate-level examination of the relationship of social policy to health and mortality in the United States, given the suggestion of promising recent work that wide health and mortality disparities in the United States may be driven by differences in social and health policies across state and local areas (Montez et al., 2020).

RECOMMENDATION 11-7: The National Institutes of Health and other public and private research entities should support a program of cross-national research aimed at understanding why trends and disparities in working-age mortality have unfolded differently in the United States and in other high-income countries. This program of research should

• examine long-terms trends and disparities, beginning in the 1950s;
• include not only transdisciplinary studies of etiology (causation pathways) but also policy research to evaluate the effectiveness of policy approaches in other countries and their potential adaptability to the United States; and
• include a complementary domestic research portfolio focused on understanding long-term changes within the United States at the state and substate levels, beginning in the 1980s when these gaps began to widen.

LESSONS FROM THE COVID-19 PANDEMIC

The COVID-19 pandemic had a devastating impact on mortality in 2020. As this report goes into print, researchers are only beginning to uncover the full impact of this virus. For many people, this pandemic marks a clear break with the past; it has changed people’s daily lives in ways previously unimaginable. In many ways, however, COVID-19 has simply reinforced and exacerbated the impact of existing social and economic inequalities within the United States. It has underscored and reinforced the importance of key themes articulated throughout this report by illustrating the ways in which economic conditions and socioeconomic inequalities make certain population groups and geographic areas more vulnerable to COVID-19.

First, this report documents increased working-age mortality from drug poisoning (Chapter 7) and cardiometabolic diseases (Chapter 9), such as hypertension, diabetes, and obesity, that during the pandemic received attention as risk factors for COVID-19 morbidity and mortality and defined vulnerable groups in need of prioritized care (Ssentongo et al., 2020; Wang et al., 2021). The increased prevalence of cardiometabolic diseases in the U.S. working-age population highlighted in this report may help explain the unexpectedly high COVID-19 death toll seen among young and middle-age adults. Moreover, there is some preliminary evidence that the stressors of the pandemic led to an increase in substance use (both alcohol and drugs) as a coping mechanism (Czeisler et al., 2020; Pollard, Tucker, and Green, 2020; United Nations Office on Drugs and Crime [UNODC], 2020), potentially foreshadowing future increases in mortality from these causes.

Second, this report examines geographic and socioeconomic disparities in health among the U.S. population, the growth in social division and income inequality, and the potential association of these disparities with trends in working-age mortality. Similarly, the pandemic exposed the heightened vulnerability of certain geographic areas (e.g., hard-hit states, rural areas, low-income neighborhoods and communities) and the economically disadvantaged to COVID-19 (Chen and Krieger, 2020; Cheng, Sun,

and Monnat, 2020; Mueller et al., 2021). Low-income individuals were disproportionately represented among service and front-line workers with the greatest exposure to the virus and were less likely than more-advantaged groups to be able to work from home, adhere to social distancing guidelines, and sustain their families (e.g., to avoid food and housing insecurity) amid a devastated economy (Weiss and Paasche-Orlow, 2020). They were also more likely to have comorbidities (Cutler, Meara, and Richards-Shubik, 2011; Pampel, Krueger, and Denney, 2010) associated with more severe COVID-19 illness (CDC, 2020a).

Third, this report documents large disparities in mortality among people of color. These disparities are reflected in the disproportionately high rates of infection, hospitalization, and death from COVID-19 experienced by Blacks and Hispanics (Ford, Reber, and Reeves, 2020; Gold et al., 2020; National Center for Health Statistics [NCHS], 2021). Although significant racial/ethnic disparities were observed at older ages, the greatest disparities occurred among younger adults (Ford, Reber, and Reeves, 2020)—so great that they exceeded overall disparities in all-cause mortality among working-age adults (NCHS, 2021). Although these racial/ethnic disparities were undoubtedly due at least in part to the geographic concentration of the initial surge in infections in large, racially and ethnically diverse central metropolitan areas, such as New York City, San Francisco, Seattle, and Los Angeles, the virus subsequently spread to less-populated and less-diverse areas of the country. By January 2021, both case and mortality rates for COVID-19 were higher in nonmetropolitan than in metropolitan counties (Ullrich and Mueller, 2021). And even in nonmetropolitan areas, large racial/ethnic disparities persisted (Cheng, Sun, and Monnat, 2020; Ford, Reber, and Reeves, 2020).

Finally, this report points to the role of health care—both access to health insurance and providers and the barriers to care delivery faced by underserved patients—in shaping progress and setbacks in working-age mortality. The pandemic and the particular difficulties experienced by the U.S. health care system (and public health infrastructure) in comparison with peer countries (Bilinski and Emanuel, 2020) underscore the barriers the public (especially marginalized groups) faced in accessing care (from COVID-19 testing, to vaccination, to intensive care) and the limited capacity of the care delivery system to absorb surges in hospitalization and other health care demands.

Thus, the COVID-19 pandemic has drawn attention to long-standing social and economic inequalities that leave some populations vulnerable when new health threats emerge. It has also highlighted the important role that public policy can play in achieving health equity. The public witnessed vivid illustrations of not only how policy decisions affected the nation’s epidemic curve but also how state trends were influenced by the decisions

of governors, legislators, and state courts, including states’ preemption of the ability of cities to enact their own mandates to prevent the spread of the virus (Haddow et al., 2020; Treskon and Docter, 2020; Wagner, Rainwater, and Carter, 2020). All these issues received greater visibility during the COVID-19 pandemic and inspired calls for policy action to address them.

CONCLUSION

The United States is losing far too many lives far too early. The rise in working-age mortality documented in this report represents a crisis, one that threatens the future of the nation’s families, workforce, economy, and national security, and one that requires action even if the evidence is imperfect or only suggestive of causal effects and solutions. This chapter has offered policy conclusions and recommendations toward addressing this crisis. At the same time, given the potential for unintended consequences of even the best-intended policy actions, it is also crucial to design policies carefully to account for potential risks, continue to monitor outcomes, generate better evidence, and adjust policies over time. The research implications of this report, also highlighted in this chapter, provide direction to this end.

TABLE 11-1 Recommendations and Policy Conclusions

This page intentionally left blank.