Raj Chetty et al.
Journal of the American Medical Association, forthcoming
Design and Setting: Income data for the US population were obtained from 1.4 billion deidentified tax records between 1999 and 2014. Mortality data were obtained from Social Security Administration death records. These data were used to estimate race- and ethnicity-adjusted life expectancy at 40 years of age by household income percentile, sex, and geographic area, and to evaluate factors associated with differences in life expectancy.
Main Outcomes and Measures: Relationship between income and life expectancy; trends in life expectancy by income group; geographic variation in life expectancy levels and trends by income group; and factors associated with differences in life expectancy across areas.
Results: The sample consisted of 1 408 287 218 person-year observations for individuals aged 40 to 76 years (mean age, 53.0 years; median household earnings among working individuals, $61 175 per year). There were 4 114 380 deaths among men (mortality rate, 596.3 per 100 000) and 2 694 808 deaths among women (mortality rate, 375.1 per 100 000). The analysis yielded 4 results. First, higher income was associated with greater longevity throughout the income distribution. The gap in life expectancy between the richest 1% and poorest 1% of individuals was 14.6 years (95% CI, 14.4 to 14.8 years) for men and 10.1 years (95% CI, 9.9 to 10.3 years) for women. Second, inequality in life expectancy increased over time. Between 2001 and 2014, life expectancy increased by 2.34 years for men and 2.91 years for women in the top 5% of the income distribution, but by only 0.32 years for men and 0.04 years for women in the bottom 5% (P < .001 for the differences for both sexes). Third, life expectancy for low-income individuals varied substantially across local areas. In the bottom income quartile, life expectancy differed by approximately 4.5 years between areas with the highest and lowest longevity. Changes in life expectancy between 2001 and 2014 ranged from gains of more than 4 years to losses of more than 2 years across areas. Fourth, geographic differences in life expectancy for individuals in the lowest income quartile were significantly correlated with health behaviors such as smoking (r = −0.69, P < .001), but were not significantly correlated with access to medical care, physical environmental factors, income inequality, or labor market conditions. Life expectancy for low-income individuals was positively correlated with the local area fraction of immigrants (r = 0.72, P < .001), fraction of college graduates (r = 0.42, P < .001), and government expenditures (r = 0.57, P < .001).
Conclusions and Relevance: In the United States between 2001 and 2014, higher income was associated with greater longevity, and differences in life expectancy across income groups increased over time. However, the association between life expectancy and income varied substantially across areas; differences in longevity across income groups decreased in some areas and increased in others. The differences in life expectancy were correlated with health behaviors and local area characteristics.
Preventive Medicine, March 2016, Pages 62–68
Methods: Data on 431,637 adults aged 30–74 and 375,354 adults aged 20–44 in the 48 contiguous US states were used from the National Longitudinal Mortality Study to estimate the impacts of state and local spending and income inequality on individual risks of all-cause and cause-specific mortality for leading causes of death in younger and middle-aged adults and older adults. To reduce bias, models incorporated state fixed effects and instrumental variables.
Results: Each additional $250 per capita per year spent on welfare predicted a 3-percentage point (− 0.031, 95% CI: − 0.059, − 0.0027) lower probability of dying from any cause. Each additional $250 per capita spent on welfare and education predicted 1.6-percentage point (− 0.016, 95% CI: − 0.031, − 0.0011) and 0.8-percentage point (− 0.008, 95% CI: − 0.0156, − 0.00024) lower probabilities of dying from coronary heart disease (CHD), respectively. No associations were found for colon cancer or chronic obstructive pulmonary disease; for diabetes, external injury, and suicide, estimates were inverse but modest in magnitude. A 0.1 higher Gini coefficient (higher income inequality) predicted 1-percentage point (0.010, 95% CI: 0.0026, 0.0180) and 0.2-percentage point (0.002, 95% CI: 0.001, 0.002) higher probabilities of dying from CHD and suicide, respectively.
Conclusions: Empirical linkages were identified between state-level spending on welfare and education and lower individual risks of dying, particularly from CHD and all causes combined. State-level income inequality predicted higher risks of dying from CHD and suicide.
Chenkai Wu et al.
Journal of Epidemiology & Community Health, forthcoming
Methods: On the basis of the Health and Retirement Study, 2956 participants who were working at baseline (1992) and completely retired during the follow-up period from 1992 to 2010 were included. Healthy retirees (n=1934) were defined as individuals who self-reported health was not an important reason to retire. The association of retirement age with all-cause mortality was analysed using the Cox model. Sociodemographic effect modifiers of the relation were examined.
Results: Over the study period, 234 healthy and 262 unhealthy retirees died. Among healthy retirees, a 1-year older age at retirement was associated with an 11% lower risk of all-cause mortality (95% CI 8% to 15%), independent of a wide range of sociodemographic, lifestyle and health confounders. Similarly, unhealthy retirees (n=1022) had a lower all-cause mortality risk when retiring later (HR 0.91, 95% CI 0.88 to 0.94). None of the sociodemographic factors were found to modify the association of retirement age with all-cause mortality.
Conclusions: Early retirement may be a risk factor for mortality and prolonged working life may provide survival benefits among US adults.
Demography, April 2016, Pages 269-293
The educational gradient in life expectancy is well documented in the United States and in other low-mortality countries. Highly educated Americans, on average, live longer than their low-educated counterparts, who have recently seen declines in adult life expectancy. However, limiting the discussion on lifespan inequality to mean differences alone overlooks other dimensions of inequality and particularly disparities in lifespan variation. The latter represents a unique form of inequality, with higher variation translating into greater uncertainty in the time of death from an individual standpoint, and higher group heterogeneity from a population perspective. Using data from the National Vital Statistics System from 1990 to 2010, this is the first study to document trends in both life expectancy and S25 — the standard deviation of age at death above 25 — by educational attainment. Among low-educated whites, adult life expectancy declined by 3.1 years for women and by 0.6 years for men. At the same time, S25 increased by about 1.5 years among high school–educated whites of both genders, becoming an increasingly important component of total lifespan inequality. By contrast, college-educated whites benefited from rising life expectancy and record low variation in age at death, consistent with the shifting mortality scenario. Among blacks, adult life expectancy increased, and S25 plateaued or declined in nearly all educational attainment groups, although blacks generally lagged behind whites of the same gender on both measures. Documenting trends in lifespan variation can therefore improve our understanding of lifespan inequality and point to diverging trajectories in adult mortality across socioeconomic strata.
American Journal of Human Biology, forthcoming
Objectives: Birth season is related to a variety of later outcomes. Among them, mortality is of great interest because it represents lifetime health outcomes. We examined the relationship between birth season and mortality in the US.
Methods: We merged the US National Health Interview Survey (NHIS) and NHIS public-use linked mortality files and analyzed 17,082 men and 19,075 women who were followed for 20 years from 1986 to 2006. We used the Cox proportional hazards model to relate birth quarter to mortality, controlling for birth year fixed effects.
Results: After controlling for years of schooling and birth year fixed effects, we found that, relative to men born in the first quarter, men born in the fourth quarter were 11% less likely to die. For women, the benefit was the largest for women born in the third quarter who were 14% less likely to die than women born in the first quarter. In the relationship between birth season and mortality, cardiovascular diseases played a noticeable role for men and malignant neoplasms for women.
Conclusions: These results were consistent with those for some developed countries, but not entirely with those for contemporary developing countries and developed countries of the past. Simple mechanisms based on the perinatal environment cannot account for the inconsistent results. We suggest that family background may play some, but not an exhaustive, role in the relationship between birth season and mortality.
Social Science & Medicine, forthcoming
Despite eliminating incidences of many diseases in the United States, parents are increasingly rejecting vaccines for their children. This article examines the reasons parents offer for doing so. It argues that parents construct a dichotomy between the natural and the artificial, in which vaccines come to be seen as unnecessary, ineffective, and potentially dangerous. Using qualitative data from interviews and observations, this article shows first, how parents view their children’s bodies, particularly from experiences of birth and with infants, as naturally perfect and in need of protection. Second, parents see vaccines as an artificial intervention that enters the body unnaturally, through injection. Third, parents perceive immunity occurring from illness to be natural and superior and immunity derived from vaccines as inferior and potentially dangerous. Finally, parents highlight the ways their own natural living serves to enhance their children’s immunity rendering vaccines unnecessary. Taken together, this dichotomy allows parents to justify rejection of vaccines as a form of protecting children’s health. These findings expose perceptions of science, technology, health, and the meanings of the body in ways that can inform public health efforts.
Laura Scherer et al.
Vaccine refusal has an impact on public health, and the human pappillomavirus (HPV) vaccine is particularly underutilized. Research suggests that it may be difficult to change vaccine-related attitudes, and there is currently no good evidence to recommend any particular intervention strategy. One reason for vaccine hesitancy is lack of trust that vaccine harms are adequately documented and reported, yet few communication strategies have explicitly attempted to improve this trust. This study tested the possibility that data from the vaccine adverse event reporting system (VAERS) can be used to increase trust that vaccine harms are adequately researched and that potential harms are disclosed to the public, and thereby improve perceptions of vaccines. In the study, participants were randomly assigned to one of three communication interventions. All participants read the Centers for Disease Control (CDC) vaccine information statement (VIS) for the HPV vaccine. Two other groups were exposed to additional information about VAERS, either summary data or full detailed reports of serious adverse events from 2013. Results showed that the CDC's VIS alone significantly increased perceptions of vaccine benefits and decreased perceived risks. Participants who were also educated about VAERS and given summary data about the serious adverse events displayed more trust in the CDC and greater HPV vaccine acceptance relative to the VIS alone. However, exposure to the detailed VAERS reports significantly reduced trust in the CDC and vaccine acceptance. Hence, general information about the VAERS data slightly increased trust in the CDC and improved vaccine acceptance, but the specific VAERS reports negatively influenced both trust and acceptance. Implications for communicating about vaccines are discussed.
Ashish Kulkarni et al.
Proceedings of the National Academy of Sciences, forthcoming
The ability to monitor the efficacy of an anticancer treatment in real time can have a critical effect on the outcome. Currently, clinical readouts of efficacy rely on indirect or anatomic measurements, which occur over prolonged time scales postchemotherapy or postimmunotherapy and may not be concordant with the actual effect. Here we describe the biology-inspired engineering of a simple 2-in-1 reporter nanoparticle that not only delivers a cytotoxic or an immunotherapy payload to the tumor but also reports back on the efficacy in real time. The reporter nanoparticles are engineered from a novel two-staged stimuli-responsive polymeric material with an optimal ratio of an enzyme-cleavable drug or immunotherapy (effector elements) and a drug function-activatable reporter element. The spatiotemporally constrained delivery of the effector and the reporter elements in a single nanoparticle produces maximum signal enhancement due to the availability of the reporter element in the same cell as the drug, thereby effectively capturing the temporal apoptosis process. Using chemotherapy-sensitive and chemotherapy-resistant tumors in vivo, we show that the reporter nanoparticles can provide a real-time noninvasive readout of tumor response to chemotherapy. The reporter nanoparticle can also monitor the efficacy of immune checkpoint inhibition in melanoma. The self-reporting capability, for the first time to our knowledge, captures an anticancer nanoparticle in action in vivo.
Anup Malani & Tomas Philipson
University of Chicago Working Paper, March 2016
Medical research and development (R&D) differs from other R&D because of a unique linkage between output and input markets for medical products: potential consumers of existing medical products are also potential subjects in clinical trials required to develop new products. Therefore, an increase in the quality or reduction in the price of an existing treatment reduces the incentive of patients to participate in trials of new treatments. We provide evidence of this linkage, which we label the “subject supply effect,” by showing that a breakthrough HIV/AIDs treatment led to a sharp drop in the supply of trial subjects after the introduction of the treatment in 1996. The subject supply effect has important positive implications for how policies such as recent insurance expansions affect the rate of medical R&D and normative implications for whether subjects ought to be compensated for enrolling in clinical trials, an ethically controversial practice.
Jamie Lynch & Paul von Hippel
Social Science & Medicine, April 2016, Pages 18–27
There is a positive gradient associating educational attainment with health, yet the explanation for this gradient is not clear. Does higher education improve health (causation)? Do the healthy become highly educated (selection)? Or do good health and high educational attainment both result from advantages established early in the life course (confounding)? This study evaluates these competing explanations by tracking changes in educational attainment and Self-rated Health (SRH) from age 15 to age 31 in the National Longitudinal Study of Youth, 1997 cohort. Ordinal logistic regression confirms that high-SRH adolescents are more likely to become highly educated. This is partly because adolescent SRH is associated with early advantages including adolescents’ academic performance, college plans, and family background (confounding); however, net of these confounders adolescent SRH still predicts adult educational attainment (selection). Fixed-effects longitudinal regression shows that educational attainment has little causal effect on SRH at age 31. Completion of a high school diploma or associate’s degree has no effect on SRH, while completion of a bachelor’s or graduate degree have effects that, though significant, are quite small (less than 0.1 points on a 5-point scale). While it is possible that educational attainment would have greater effect on health at older ages, at age 31 what we see is a health gradient in education, shaped primarily by selection and confounding rather than by a causal effect of education on health.
Riley Steiner et al.
JAMA Pediatrics, forthcoming
Importance: Long-acting reversible contraception (LARC), specifically intrauterine devices and implants, offers an unprecedented opportunity to reduce unintended pregnancies among adolescents because it is highly effective even with typical use. However, adolescent LARC users may be less likely to use condoms for preventing sexually transmitted infections compared with users of moderately effective contraceptive methods (ie, oral, Depo-Provera injection, patch, and ring contraceptives).
Design, Setting, and Participants: Cross-sectional analysis using data from the 2013 national Youth Risk Behavior Survey, a nationally representative sample of US high school students in grades 9 through 12. Descriptive analyses were conducted among sexually active female students (n = 2288); logistic regression analyses were restricted to sexually active female users of LARC and moderately effective contraception (n = 619). The analyses were conducted in July and August 2015.
Results: Among the 2288 sexually active female participants (56.7% white; 33.6% in 12th grade), 1.8% used LARC; 5.7% used Depo-Provera, patch, or ring; 22.4% used oral contraceptives; 40.8% used condoms; 11.8% used withdrawal or other method; 15.7% used no contraceptive method; and 1.9% were not sure. In adjusted analyses, LARC users were about 60% less likely to use condoms compared with oral contraceptive users (adjusted prevalence ratio [aPR], 0.42; 95% CI, 0.21-0.84). No significant differences in condom use were observed between LARC users and Depo-Provera injection, patch, or ring users (aPR, 0.57; 95% CI, 0.26-1.25). The LARC users were more than twice as likely to have 2 or more recent sexual partners compared with oral contraceptive users (aPR, 2.61; 95% CI, 1.75-3.90) and Depo-Provera, patch, or ring users (aPR, 2.58; 95% CI, 1.17-5.67).
Conclusions and Relevance: Observed differences in condom use may reflect motivations to use condoms for backup pregnancy prevention. Users of highly effective LARC methods may no longer perceive a need for condoms even if they have multiple sexual partners, which places them at risk for sexually transmitted infections. As uptake of LARC increases among adolescents, a clear need exists to incorporate messages about condom use specifically for sexually transmitted infection prevention.
Troy Quast & Fidel Gonzalez
Health Economics, forthcoming
While reducing the transmission of sexually transmitted infections is a common argument for regulating sex work, relatively little empirical evidence is available regarding the effectiveness of these policies. We investigate the effects of highly publicized sex work regulations introduced in 2005 in Tijuana, Mexico on the incidence of trichomoniasis. State-level, annual data for the 1995–2012 period are employed that include the incidence rates of trichomoniasis by age group and predictor variables. We find that the regulations led to a decrease in the incidence rate of trichomoniasis. Specifically, while our estimates are somewhat noisy, the all-ages incidence rate in the 2005–2012 period is roughly 37% lower than what is predicted by our synthetic control estimates and corresponds to approximately 800 fewer reported cases of trichomoniasis per year. We find that the decreases are especially pronounced for 15–24 and 25–44 age cohorts.
Holly Witteman et al.
Health Affairs, April 2016, Pages 726-733
As people increasingly turn to social media to access and create health evidence, the greater availability of data and information ought to help more people make evidence-informed health decisions that align with what matters to them. However, questions remain as to whether people can be swayed in favor of or against options by polarized social media, particularly in the case of controversial topics. We created a composite mock news article about home birth from six real news articles and randomly assigned participants in an online study to view comments posted about the original six articles. We found that exposure to one-sided social media comments with one-sided opinions influenced participants’ opinions of the health topic regardless of their reported level of previous knowledge, especially when comments contained personal stories. Comments representing a breadth of views did not influence opinions, which suggests that while exposure to one-sided comments may bias opinions, exposure to balanced comments may avoid such bias.
Carl Bonander, Anders Jonsson & Finn Nilson
European Journal of Public Health, April 2016, Pages 334-338
Background: Annually, 100 people die as a result of residential fires in Sweden and almost a third of the fatal fires are known to be caused by smoking. In an attempt to reduce the occurrence of these events, reduced ignition propensity (RIP) cigarettes have been developed. They are designed to reduce the risk of fire by preventing the cigarette from burning through the full length when left unattended. In November 2011, a ban was introduced, forbidding the production and sale of all non-RIP cigarettes in all member states of the European Union, including Sweden.
Methods: Monthly data on all recorded residential fires and associated fatalities in Sweden from January 2000 to December 2013 were analyzed using an interrupted time series design. The effect of the intervention [in relative risk (RR)] was quantified using generalised additive models for location, shape and scale.
Results: There were no statistically significant intervention effects on residential fires (RR 0.95 [95% CI: 0.89–1.01]), fatal residential fires (RR 0.99 [95% CI: 0.80–1.23]), residential fires where smoking was a known cause (RR 1.10 [95% CI: 0.95–1.28]) or fatal residential fires where smoking was a known cause (RR 0.92 [95% CI: 0.63–1.35]).
Conclusion: No evidence of an effect of the ban on all non-RIP cigarettes on the risk of residential fires in Sweden was found. The results may not be generalisable to other countries.