Advertisement
Research Article| Volume 64, ISSUE 4, SUPPLEMENT 1, S53-S62, April 2023

Mortality Relative Risks by Smoking, Race/Ethnicity, and Education

Open AccessPublished:February 10, 2023DOI:https://doi.org/10.1016/j.amepre.2022.12.006

      Introduction

      The impact of cigarette smoking on mortality is well studied, with estimates of the relative mortality risks for the overall population widely available. However, age-specific mortality estimates for different sociodemographic groups in the U.S. are lacking.

      Methods

      Using the 1987–2018 National Health Interview Survey Linked Mortality Files through 2019, all-cause mortality relative risks (RRs) were estimated for current smokers or recent quitters and long-term quitters compared with those for never smokers. Stratified Cox proportional hazards regression models were used to estimate RRs by age, gender, race/ethnicity, and educational attainment. RRs were also assessed for current smokers or recent quitters by smoking intensity and for long-term quitters by years since quitting. The analysis was conducted in 2021–2022.

      Results

      All-cause mortality RRs among current smokers or recent quitters were generally highest for non-Hispanic White individuals than for never smokers, followed by non-Hispanic Black individuals, and were lowest for Hispanic individuals. RRs varied greatly by educational attainment; generally, higher-education groups had greater RRs associated with smoking than lower-education groups. Conversely, the RRs by years since quitting among long-term quitters did not show clear differences across race/ethnicity and education groups. Age-specific RR patterns varied greatly across racial/ethnic and education groups as well as by gender.

      Conclusions

      Age-specific all-cause mortality rates associated with smoking vary considerably by sociodemographic factors. Among high-education groups, lower underlying mortality rates for never smokers result in correspondingly high RR estimates for current smoking. These estimates can be incorporated in modeling analyses to assess tobacco control interventions’ impact on smoking-related health disparities between different sociodemographic groups.

      INTRODUCTION

      Cigarette smoking has declined significantly in the U.S., yet the extent of the decrease varies by sociodemographic factors,
      HHS
      The health consequences of smoking-50 years of progress: a report of the Surgeon General.
      ,
      U.S. National Cancer Institute
      A socioecological approach to addressing tobacco-related health disparities. National Cancer Institute Tobacco Control Monograph 22.
      resulting in persistent disparities between different sociodemographic groups.
      HHS
      The health consequences of smoking-50 years of progress: a report of the Surgeon General.
      U.S. National Cancer Institute
      A socioecological approach to addressing tobacco-related health disparities. National Cancer Institute Tobacco Control Monograph 22.
      • Jamal A
      • Homa DM
      • O'Connor E
      • et al.
      Current cigarette smoking among adults - United States, 2005-2014.
      • Kanjilal S
      • Gregg EW
      • Cheng YJ
      • et al.
      Socioeconomic status and trends in disparities in 4 major risk factors for cardiovascular disease among U.S. adults, 1971–2002.
      • Meza R
      • Cao P
      • Jeon J
      • et al.
      Patterns of birth cohort-specific smoking histories by race and ethnicity in the U.S.
      • Cao P
      • Jeon J
      • Tam J
      • et al.
      Smoking disparities by level of educational attainment and birth cohort in the U.S.
      • Jeon J
      • Cao P
      • Fleischer NL
      • et al.
      Birth cohort-specific smoking patterns by family income in the U.S.
      Smoking behaviors and related mortality vary by several factors, including age, gender, race/ethnicity, SES, and region.
      • Ho JY
      • Elo IT
      The contribution of smoking to black-white differences in U.S. Mortality.
      • Lariscy JT
      • Hummer RA
      • Hayward MD
      Hispanic older adult mortality in the United States: new estimates and an assessment of factors shaping the Hispanic paradox.
      • Lariscy JT
      • Hummer RA
      • Rogers RG
      Cigarette smoking and all-cause and cause-specific adult mortality in the United States.
      • Fenelon A
      • Preston SH
      Estimating smoking-attributable mortality in the United States.
      • Holford TR
      • Levy DT
      • Meza R
      Comparison of smoking history patterns among African American and white cohorts in the United States born 1890 to 1990.
      • Ho JY
      • Fenelon A
      The contribution of smoking to educational gradients in U.S. Life expectancy.
      Cigarette smoking is estimated to cause >480,000 deaths each year in the U.S.,
      HHS
      The health consequences of smoking-50 years of progress: a report of the Surgeon General.
      ,

      Tobacco-related mortality. Centers for Disease Control and Prevention.https://www.cdc.gov/tobacco/data_statistics/fact_sheets/health_effects/tobacco_related_mortality/index.htm. Accessed January 18, 2023.

      but this mortality burden is unequally distributed across the nation. People with lower levels of educational attainment are known to smoke at higher rates than those with more education,
      • Cao P
      • Jeon J
      • Tam J
      • et al.
      Smoking disparities by level of educational attainment and birth cohort in the U.S.
      ,
      • Ho JY
      • Fenelon A
      The contribution of smoking to educational gradients in U.S. Life expectancy.
      ,
      • Pampel FC
      The persistence of educational disparities in smoking.
      • Maralani V
      Understanding the links between education and smoking.
      • Zhu BP
      • Giovino GA
      • Mowery PD
      • Eriksen MP
      The relationship between cigarette smoking and education revisited: implications for categorizing persons’ educational status.
      but the smoking-related mortality rate they face has not been well studied. Mortality rates associated with smoking likely vary by race/ethnicity as well,
      • Ho JY
      • Elo IT
      The contribution of smoking to black-white differences in U.S. Mortality.
      ,
      • Rostron BL
      • Lynn BCD
      • Chang CM
      • Ren C
      • Salazar E
      • Ambrose BK
      The contribution of smoking-attributable mortality to differences in mortality and life expectancy among U.S. African-American and white adults, 2000–2019.
      although this is also an understudied topic. One recent study by Inoue-Choi et al.
      • Inoue-Choi M
      • Christensen CH
      • Rostron BL
      • et al.
      Dose-response association of low-intensity and nondaily smoking with mortality in the United States.
      found that the RR of all-cause mortality for current daily smoking compared with that for never smoking was higher in non-Hispanic White individuals than in non-Hispanic Black or Hispanic individuals. Mortality rates are also known to vary by smoking intensity, measured by the number of cigarettes smoked per day or month among current smokers and by time since quitting among former smokers, though this is also understudied for specific subpopulations.
      • Ho JY
      • Elo IT
      The contribution of smoking to black-white differences in U.S. Mortality.
      ,
      • Lariscy JT
      • Hummer RA
      • Rogers RG
      Cigarette smoking and all-cause and cause-specific adult mortality in the United States.
      ,
      • Inoue-Choi M
      • Christensen CH
      • Rostron BL
      • et al.
      Dose-response association of low-intensity and nondaily smoking with mortality in the United States.
      Empirical estimates of mortality rates associated with current and former smoking are necessary both to quantify the impact of smoking on mortality in a given population and to make projections about the future burden of smoking—typically assessed using tobacco simulation models.
      • Moolgavkar SH
      • Holford TR
      • Levy DT
      • et al.
      Impact of reduced tobacco smoking on lung cancer mortality in the United States during 1975–2000.
      • Tam J
      • Levy DT
      • Jeon J
      • et al.
      Projecting the effects of tobacco control policies in the USA through microsimulation: a study protocol.
      • Jeon J
      • Holford TR
      • Levy DT
      • et al.
      Smoking and lung cancer mortality in the United States from 2015 to 2065: a comparative modeling approach.
      Although previous research has produced detailed age- and gender-specific all-cause mortality RRs for the general U.S. population,
      • Rosenberg MA
      • Feuer EJ
      • Yu B
      • et al.
      Chapter 3: cohort life tables by smoking status, removing lung cancer as a cause of death.
      such estimates are comparatively absent for specific sociodemographic groups—an impediment to future research that aims to both quantify and monitor smoking-related disparities over time. Recent data for the general U.S. population suggest that the risks of mortality from cigarette smoking relative to never smoking may be increasing.
      HHS
      The health consequences of smoking-50 years of progress: a report of the Surgeon General.
      ,
      • Thun MJ
      • Carter BD
      • Feskanich D
      • et al.
      50-year trends in smoking-related mortality in the United States.
      This further shows the need for up-to-date analyses of the mortality impacts of cigarette smoking across the whole U.S. population, in addition to specific subgroups.
      Assessing the long-term consequences of cigarette smoking in specific populations requires detailed estimates of their smoking-related health risks by age and smoking history. However, as discussed earlier, current estimates of smoking-related mortality RRs in the U.S. do not have the required details to inform simulation models of the health consequences of smoking for relevant sociodemographic groups. To address these gaps, nationally representative data were used to estimate age-specific all-cause mortality RRs for current smokers or recent quitters who quit smoking <2 years ago and long-term quitters who quit smoking at least 2 years ago compared with those of never smokers, stratified by gender, race/ethnicity, and educational attainment.

      METHODS

      Study Sample

      The National Health Interview Survey (NHIS) collects annual data on demographic characteristics and health information of the civilian non-institutionalized population of the U.S. Adult tobacco use data have been collected since 1965. On the basis of the availability and consistency of smoking data/questions, including use frequency and intensity, data from 1987, 1988, 1990, 1994, 1995, and 1997–2018 NHIS surveys linked with mortality data from the 2019 National Death Index were utilized.

      NCHS data linkage. NCHS data linked to NDI mortality files. Centers for Disease Control and Prevention.https://www.cdc.gov/nchs/data-linkage/mortality.htm?CDC_AA_refVal=https%3A%2F%2Fwww.cdc.gov%2Fnchs%2Fdata_access%2Fdata_linkage%2Fmortality.htm. Accessed January 18, 2023.

      LA Blewett, JA Rivera Drew, ML King, KCW Williams, N Del Ponte, P Convey, IPUMS Health Surveys: National Health Interview Survey, version 7.2. Minneapolis, MN: IPUMS. https://doi.org/10.18128/D070.V7.2. Published 2022.

      National Center for Health StatisticsCenters for Disease Control and Prevention
      The linkage of National Center for Health Statistics survey data to the National Death Index — 2019 linked mortality file (LMF): linkage methodology and analytic considerations.
      Covariates included age; gender; race/ethnicity; education; and cigarette smoking variables such as use status (never, current, former), smoking intensity measured by the average number of cigarettes smoked per day (CPD), and years since quitting (YSQ) for former smokers. People who were not eligible for mortality follow-up (27,215) and missing information for smoking status (20,434), race/ethnicity (583), or education (2,862) were excluded. A total of 611,275 participants aged 30–84 years were included in the race/ethnicity analysis, and 608,996 participants were included in the education analysis. Individuals were followed for all-cause mortality for 10 years of follow-up from their survey interview; a total of 27,976 and 26,662 deaths were observed for males and females in the race/ethnicity analysis, respectively, and a total of 27,813 and 26,523 deaths were observed for males and females in the education analysis, respectively.

      Measures

      Ever smokers were defined as individuals who smoked 100 cigarettes or more during their lifetime. They were classified as current smokers if they reported smoking every day or some days and former smokers if they reported not smoking at all at the time of the survey interview. Former smokers were categorized into 2 groups: recent quitters (i.e., those who quit smoking <2 years ago) versus long-term quitters (i.e., those who quit smoking 2 or more years ago). In the main analysis, current smokers and recent quitters were combined into a single group because recent quitters are (1) much more likely to relapse back to smoking
      • Herd N
      • Borland R
      • Hyland A.
      Predictors of smoking relapse by duration of abstinence: findings from the International Tobacco Control (ITC) Four Country Survey.
      • Hughes JR
      • Peters EN
      • Naud S.
      Relapse to smoking after 1 year of abstinence: a meta-analysis.
      • Krall EA
      • Garvey AJ
      • Garcia RI.
      Smoking relapse after 2 years of abstinence: findings from the VA Normative Aging Study.
      and (2) more likely to have quit as a response to smoking-related disease or disability.
      • Kalkhoran S
      • Kruse GR
      • Chang Y
      • Rigotti NA.
      Smoking Cessation efforts by U.S. adult smokers with medical comorbidities.
      • Fry JS
      • Lee PN
      • Forey BA
      • Coombs KJ.
      How rapidly does the excess risk of lung cancer decline following quitting smoking? A quantitative review using the negative exponential model.
      • Knoke JD
      • Burns DM
      • Thun MJ.
      The change in excess risk of lung cancer attributable to smoking following smoking cessation: an examination of different analytic approaches using CPS-I data.
      Thus, their mortality rate is inherently higher than that of an established former smoker. Age was categorized into 5 groups: 30–44, 45–54, 55–64, 65–74, and 75–84 years. Race/ethnicity was categorized into 5 groups: non-Hispanic White (NHW), non-Hispanic Black (NHB), and Hispanic. Non-Hispanic individuals of other racial backgrounds were excluded from the race/ethnicity-specific analyses owing to insufficient sample sizes. Education was categorized into 5 groups: ≤8th grade, 9th–11th grade, high school graduate or GED, some college, and at least a college degree.

      Statistical Analysis

      RRs of all-cause mortality, measured as Cox hazard ratios, and their 95% CIs were estimated using Cox proportional hazards models, accounting for the complex sample design of the NHIS using Taylor series linearization. Gender-specific RRs were estimated for current smokers or recent quitters versus never smokers and for long-term quitters versus never smokers. Separate stratified analyses were conducted by age and race/ethnicity and by age and educational attainment as an SES indicator.
      First, RRs for all-cause mortality were estimated for current smokers or recent quitters and long-term quitters versus never smokers in the U.S. population. Further analyses were then conducted splitting the current smokers or recent quitters by smoking intensity (<10, 10–19, 20, ≥21 CPD or 1–5, 6–15, 16–25, ≥26 CPD) and the long-term quitters by the length of abstinence (2 to <5, 5 to <15, ≥15 YSQ). Analyses were conducted using the SURVEYPHREG procedure in SAS software, Version 9.4.

      RESULTS

      Table 1 summarizes the demographic characteristics of the study sample, stratified by smoking status: never smoker, current smoker, former smoker (YSQ<2, 2≤YSQ<5, 5≤YSQ<15, YSQ≥15). About 40% or above were in the age group 30–44 years in all smoking categories, except for former users who quit smoking 15 or more years ago. This group of former users was considerably older than those in other smoking status groups. A higher percentage of females than males were never smokers. The proportion of NHB individuals was higher in current smokers than in former smokers. The proportion of individuals with at least a college degree was lowest in current smokers and highest in never smokers, increasing with a longer length of smoking cessation among former smokers. Conversely, the percentage of individuals with 9th–11th-grade education or high school education or GED was highest in current smokers and lowest in never smokers, with decreasing proportion by more YSQ in former users.
      Table 1Demographic Characteristics in Different Cigarette Use Groups in 1987–2018 NHIS Data
      Former smoker
      CharacteristicsNever smoker (n=330,500),Current smoker (n=133,448),YSQ<2 (n=4,337),2≤YSQ<5 (n=16,895),5≤YSQ<15 (n=44,674),YSQ≥15 (n=82,004),
      % (95% CI)% (95% CI)% (95% CI)% (95% CI)% (95% CI)% (95% CI)
      Age group, years
       30–4441.3 (41.0, 41.6)44.6 (44.3, 45.0)46.9 (45.0, 48.7)43.0 (42.0, 43.9)38.0 (37.5, 38.6)10.0 (9.7, 10.3)
       45–5423.1 (22.9, 23.3)27.2 (26.9, 27.5)21.3 (19.6, 23.0)22.3 (21.5, 23.1)22.4 (21.9, 22.9)20.1 (19.8, 20.5)
       55–6416.9 (16.7, 17.1)17.6 (17.4, 17.9)17.8 (16.4, 19.3)18.6 (17.9, 19.3)19.6 (19.2, 20.1)26.3 (25.9, 26.7)
       65–7411.3 (11.1, 11.4)8.0 (7.9, 8.2)10.9 (9.9, 12.1)11.9 (11.3, 12.5)14.2 (13.8, 14.6)25.9 (25.6, 26.3)
       75–847.4 (7.3, 7.5)2.5 (2.4, 2.6)3.1 (2.5, 3.8)4.3 (3.9, 4.6)5.7 (5.5, 6.0)17.6 (17.2, 17.9)
      Sex
       Male41.5 (41.3, 41.8)53.0 (52.6, 53.3)55.2 (53.4, 57.0)54.5 (53.5, 55.5)55.1 (54.5, 55.7)58.4 (57.9, 58.8)
       Female58.5 (58.2, 58.7)47.0 (46.7, 47.4)44.8 (43.0, 46.6)45.5 (44.5, 46.5)44.9 (44.3, 45.5)41.6 (41.2, 42.1)
      Race/ethnicity
       Non-Hispanic White67.2 (66.7, 67.6)76.1 (75.6, 76.5)79.8 (78.3, 81.1)77.4 (76.6, 78.3)79.1 (78.6, 79.7)84.2 (83.8, 84.6)
       Non-Hispanic Black12.1 (11.8, 12.5)12.3 (11.9, 12.7)9.1 (8.2, 10.2)9.0 (8.5, 9.5)8.0 (7.7, 8.4)6.5 (6.2, 6.7)
       Hispanic6.5 (6.3, 6.7)3.5 (3.3, 3.7)3.2 (2.5, 4.1)3.9 (3.5, 4.3)3.7 (3.5, 4.0)2.7 (2.5, 2.9)
       Non-Hispanic other14.2 (13.9, 14.6)8.1 (7.9, 8.4)7.9 (7.0, 8.9)9.7 (9.1, 10.3)9.1 (8.8, 9.5)6.7 (6.4, 6.9)
      Education
       ≤8th grade6.8 (6.6, 7.0)6.3 (6.1, 6.5)7.4 (6.6, 8.3)5.9 (5.5, 6.4)5.9 (5.7, 6.2)6.3 (6.1, 6.5)
       9–11th grade5.6 (5.5, 5.7)12.8 (12.6, 13.0)9.8 (8.8, 10.9)8.1 (7.6, 8.6)7.6 (7.3, 7.9)6.9 (6.7, 7.1)
       HSG or GED27.2 (27.0, 27.5)40.2 (39.9, 40.6)36.4 (34.7, 38.2)33.9 (33.0, 34.8)32.1 (31.5, 32.6)29.5 (29.1, 29.9)
       Some college25.1 (24.9, 25.3)27.7 (27.4, 28.1)26.6 (24.9, 28.3)29.2 (28.4, 30.1)29.0 (28.5, 29.6)27.6 (27.2, 28.0)
       At least a college degree35.3 (34.9, 35.7)12.9 (12.6, 13.2)19.8 (18.3, 21.4)22.9 (22.0, 23.7)25.4 (24.8, 25.9)29.7 (29.2, 30.2)
      Cigarettes per day
      Average number of cigarettes smoked per day (current smokers) and average number of cigarettes usually smoked per day when they smoked regularly (former smokers). Information for average cigarettes per day for former smokers was only available in 5 years of the survey: 1987, 1988, 1990, 2005, 2010. HSG, high school graduate; NHIS, National Health Interview Survey; YSQ, years since quitting.
       <622.0 (21.7, 22.3)15.9 (14.5, 17.4)14.7 (13.3, 16.2)15.3 (14.4, 16.1)19.8 (19.1, 20.6)
       6–1532.4 (32.1, 32.8)27.3 (25.6, 29.0)26.7 (24.9, 28.5)25.5 (24.5, 26.5)26.4 (25.6, 27.2)
       16–2531.7 (31.3, 32.0)33.6 (31.8, 35.4)32.7 (30.9, 34.6)33.4 (32.3, 34.4)31.7 (30.9, 32.6)
       ≥2613.9 (13.6, 14.2)23.3 (21.8, 24.8)25.9 (24.4, 27.5)25.8 (24.9, 26.8)22.1 (21.4, 22.8)
      a Average number of cigarettes smoked per day (current smokers) and average number of cigarettes usually smoked per day when they smoked regularly (former smokers). Information for average cigarettes per day for former smokers was only available in 5 years of the survey: 1987, 1988, 1990, 2005, 2010.HSG, high school graduate; NHIS, National Health Interview Survey; YSQ, years since quitting.
      For both genders, all-cause mortality RRs for current smokers or recent quitters were highest in NHW individuals, followed by NHB individuals and then Hispanic individuals in most age groups (Table 2 and Figure 1). Mortality RRs increased until ages 55–64 years, then declined in NHW males. Conversely, RRs were relatively constant between ages 30 and 54 years, then increased and peaked at ages 65–74 years, and then decreased in ages 75–84 years in NHB males. RRs were highest at ages 65–74 years, the second highest at ages 30–44 years, and relatively constant at the other age groups in Hispanic males. RRs in NHW females showed similar age patterns as in NHS males. The RRs in NHB females are similar across all age groups. Among current smokers or recent quitters, the difference in RRs between females and males varied by age and race/ethnicity, although the differences were not statistically significant in some categories, in which 95% CIs for males and females were overlapping. In NHW individuals, the RRs were lower in males than in females for the age groups 30–44 and 75–84 years. The RRs were higher in NHB males than in females at ages 65–74 years. RRs in Hispanic individuals do not show a clear pattern by gender. The RRs for long-term quitters versus never smokers increased until ages 65–74 years and then declined in both NHW males and females, but there were no clear age patterns in NHB and Hispanic populations (Table 2 and Appendix Figure 1, available online).
      Table 2All-Cause Mortality RRs for Current Smoker or Recent Quitters and Long-Term Quitters by Race/Ethnicity
      MalesFemales
      VariablesCurrent smoker or recent quitter,
      Recent quitters are individuals who quit smoking <2 years ago.
      Long-term quitter,
      Long-term quitters are individuals who have been quitting smoking ≥2 years.
      Current smoker or recent quitter,
      Recent quitters are individuals who quit smoking <2 years ago.
      Long-term quitter,
      Long-term quitters are individuals who have been quitting smoking ≥2 years.
      Current smoker or recent quitter,
      Recent quitters are individuals who quit smoking <2 years ago.
      Long-term quitter,
      Long-term quitters are individuals who have been quitting smoking ≥2 years.
      RR (95% CI)RR (95% CI)RR (95% CI)RR (95% CI)Males/females RR ratioMales/females

      RR ratio
      All races, years
       30–442.14 (1.86, 2.46)1.16 (0.92, 1.44)2.28 (1.93, 2.70)1.09 (0.81, 1.45)0.941.06
       45–542.66 (2.37, 2.97)1.19 (1.02, 1.37)2.60 (2.31, 2.93)1.16 (0.98, 1.37)1.021.03
       55–643.38 (3.09, 3.69)1.41 (1.28, 1.56)3.20 (2.91, 3.51)1.44 (1.29, 1.61)1.060.98
       65–743.35 (3.12, 3.60)1.51 (1.41, 1.62)2.96 (2.77, 3.17)1.55 (1.45, 1.67)1.130.97
       75–842.36 (2.20, 2.53)1.40 (1.32, 1.47)2.72 (2.57, 2.88)1.47 (1.40, 1.55)0.870.95
      Non-Hispanic White, years
       30–442.29 (1.92, 2.74)1.18 (0.87, 1.58)3.10 (2.46, 3.91)1.34 (0.93, 1.93)0.740.88
       45–543.15 (2.74, 3.63)1.17 (0.96, 1.41)2.81 (2.40, 3.29)1.21 (0.98, 1.49)1.120.97
       55–643.75 (3.38, 4.16)1.51 (1.33, 1.70)3.60 (3.21, 4.04)1.51 (1.32, 1.72)1.041.00
       65–743.45 (3.16, 3.76)1.51 (1.39, 1.64)3.22 (2.98, 3.48)1.62 (1.50, 1.76)1.070.93
       75–842.43 (2.24, 2.64)1.41 (1.32, 1.50)2.82 (2.65, 3.00)1.51 (1.43, 1.59)0.860.93
      Non-Hispanic Black, years
       30–441.99 (1.43, 2.77)1.44 (0.86, 2.42)2.01 (1.40, 2.87)1.48 (0.66, 3.34)0.990.97
       45–541.84 (1.47, 2.30)1.51 (1.08, 2.09)2.31 (1.83, 2.91)1.21 (0.87, 1.68)0.801.25
       55–642.76 (2.25, 3.40)1.48 (1.18, 1.86)2.26 (1.89, 2.71)1.51 (1.17, 1.95)1.220.98
       65–743.08 (2.61, 3.63)1.60 (1.33, 1.93)2.30 (1.96, 2.69)1.39 (1.17, 1.66)1.341.15
       75–842.01 (1.68, 2.42)1.21 (1.04, 1.42)1.99 (1.67, 2.37)1.10 (0.96, 1.28)1.011.10
      Hispanic, years
       30–442.26 (1.58, 3.24)1.51 (0.92, 2.48)1.17 (0.72, 1.88)1.10 (0.54, 2.24)1.931.37
       45–541.73 (1.25, 2.37)1.34 (0.95, 1.89)1.81 (1.20, 2.75)1.35 (0.79, 2.30)0.960.99
       55–641.84 (1.36, 2.49)0.98 (0.72, 1.35)2.19 (1.52, 3.15)1.42 (0.97, 2.07)0.840.69
       65–742.58 (2.03, 3.27)1.52 (1.20, 1.92)1.56 (1.15, 2.11)1.42 (1.08, 1.87)1.651.07
       75–841.65 (1.27, 2.15)1.17 (0.94, 1.46)2.14 (1.59, 2.89)1.41 (1.12, 1.78)0.770.83
      Note: Data for all-cause mortality within a 10-year follow-up period ranges from 1987 to 2019. RR (i.e., hazard ratio) was estimated from the Cox proportional hazard regression analysis.
      a Recent quitters are individuals who quit smoking <2 years ago.
      b Long-term quitters are individuals who have been quitting smoking ≥2 years.
      Figure 1
      Figure 1RRs of all-cause mortality for current smokers or recent quitters versus for never smokers by age group and race/ethnicity (top panels) and by age group and educational attainment level (bottom panels).
      Note: These RRs are hazard ratios estimated from the Cox proportional hazard regression analysis stratified by gender (males, left panels; females, right panels). Recent quitters are individuals who quit smoking <2 years ago. All denotes all race/ethnicity combined
      HSG, high school graduate.
      Table 3 and Figure 1 show RR estimates by educational attainment. The results show in general a positive gradient in all-cause mortality RRs for current smokers or recent quitters by educational attainment level among males aged 30–44, 45–54, and 55–64 years. RRs were generally higher among lower-education groups than among higher-education groups in males aged 65–74 years, but they were lowest in the ≤8th grade and some college groups and highest in the at-least-a-college-degree group in males aged 75–84 years. The RRs patterns by educational attainment level in females differed from those in males. The RRs in females with at least a college degree were lowest in ages 30–44 years but became highest in ages 65–74 and 75–84 years. In addition, there was a clear positive gradient in all-cause mortality RRs by educational attainment level among females aged 65–74 and 75–84 years. The age pattern in RRs varied depending on educational attainment, but in general, it increased by age and peaked either at ages 55–64 or at 65–74 years and then declined in males. In contrast, RRs increased by age among females with at least a college degree, but the age pattern is less clear in the other education groups in females. Gender differences in RRs for current smokers or recent quitters varied by age and educational attainment, although differences were not statistically significant in some categories (95% CIs were overlapping). The RRs for those with at least a college degree were higher in males than in females in ages below 65 years but lower after age 65 years; they were lower in males than in females with some college education at ages 30–44 and 75–84 years. No clear RR patterns were observed by educational attainment level for long-term quitters versus never smokers (Table 3 and Appendix Figure 1, available online).
      Table 3All-Cause Mortality RRs for Current Smoker or Recent Quitters and Long-Term Quitters by Education
      MalesFemales
      VariablesCurrent smoker or recent quitter,
      Recent quitters are individuals who quit smoking <2 years ago.
      Long-term quitter,
      Long-term quitters are individuals who have been quitting smoking ≥2 years. HSG, high school graduate.
      Current smoker or recent quitter,
      Recent quitters are individuals who quit smoking <2 years ago.
      Long-term quitter,
      Long-term quitters are individuals who have been quitting smoking ≥2 years. HSG, high school graduate.
      Current smoker or recent quitter,
      Recent quitters are individuals who quit smoking <2 years ago.
      Long-term quitter,
      Long-term quitters are individuals who have been quitting smoking ≥2 years. HSG, high school graduate.
      RR (95% CI)RR (95% CI)RR (95% CI)RR (95% CI)Males/females RR ratioMales/females RR ratio
      ≤8th grade, years
       30–441.56 (0.98, 2.49)1.15 (0.53, 2.47)1.85 (1.00, 3.45)0.84
       45–541.97 (1.31, 2.97)0.88 (0.53, 1.48)2.11 (1.39, 3.22)1.61 (0.78, 3.32)0.930.55
       55–642.52 (1.92, 3.30)1.47 (1.08, 2.01)3.03 (2.34, 3.93)1.71 (1.17, 2.50)0.830.86
       65–743.22 (2.69, 3.84)1.70 (1.42, 2.03)2.46 (2.04, 2.97)1.80 (1.47, 2.22)1.310.94
       75–842.00 (1.74, 2.29)1.28 (1.15, 1.44)2.40 (2.07, 2.77)1.43 (1.26, 1.63)0.830.90
      9th–11th grade, years
       30–441.69 (1.10, 2.59)0.47 (0.20, 1.10)2.30 (1.46, 3.63)2.25 (0.90, 5.63)0.730.21
       45–541.39 (1.02, 1.89)0.66 (0.41, 1.08)2.69 (1.96, 3.70)1.70 (0.97, 2.98)0.520.39
       55–642.44 (1.85, 3.21)1.34 (0.97, 1.84)2.47 (1.94, 3.14)1.69 (1.24, 2.31)0.990.79
       65–742.95 (2.39, 3.64)1.58 (1.28, 1.95)2.71 (2.30, 3.19)1.74 (1.45, 2.09)1.090.91
       75–842.41 (1.95, 2.97)1.69 (1.42, 2.02)2.51 (2.16, 2.92)1.66 (1.46, 1.88)0.961.02
      HSG or GED, years
       30–441.78 (1.40, 2.25)0.91 (0.63, 1.33)1.76 (1.33, 2.33)1.04 (0.62, 1.75)1.010.88
       45–542.02 (1.68, 2.43)1.00 (0.79, 1.28)1.89 (1.54, 2.32)1.06 (0.80, 1.42)1.070.94
       55–642.63 (2.25, 3.07)1.24 (1.04, 1.48)2.85 (2.45, 3.30)1.52 (1.25, 1.83)0.920.82
       65–742.97 (2.60, 3.40)1.49 (1.30, 1.70)2.76 (2.46, 3.09)1.60 (1.43, 1.79)1.080.93
       75–842.34 (2.05, 2.67)1.41 (1.28, 1.54)2.60 (2.38, 2.85)1.62 (1.49, 1.75)0.900.87
      Some college, years
       30–441.65 (1.27, 2.13)1.21 (0.83, 1.78)2.44 (1.83, 3.27)1.18 (0.74, 1.86)0.681.03
       45–542.62 (2.12, 3.24)1.29 (0.99, 1.70)2.57 (2.07, 3.18)1.01 (0.74, 1.37)1.021.28
       55–643.04 (2.55, 3.64)1.25 (1.02, 1.53)3.15 (2.59, 3.83)1.33 (1.07, 1.65)0.970.94
       65–742.67 (2.28, 3.12)1.26 (1.08, 1.46)2.73 (2.35, 3.18)1.41 (1.20, 1.65)0.980.89
       75–842.00 (1.66, 2.41)1.28 (1.13, 1.45)2.96 (2.57, 3.40)1.42 (1.27, 1.60)0.680.90
      At least a college degree, years
       30–442.09 (1.44, 3.03)1.52 (0.93, 2.48)1.52 (0.91, 2.54)1.02 (0.57, 1.80)1.381.49
       45–543.06 (2.30, 4.07)1.51 (1.12, 2.04)2.32 (1.58, 3.40)1.32 (0.92, 1.89)1.321.14
       55–643.54 (2.79, 4.50)1.42 (1.16, 1.74)2.73 (2.07, 3.60)1.38 (1.08, 1.77)1.301.03
       65–743.06 (2.57, 3.65)1.33 (1.14, 1.54)3.78 (3.01, 4.75)1.58 (1.25, 2.00)0.810.84
       75–842.56 (2.08, 3.15)1.27 (1.12, 1.44)3.76 (3.11, 4.55)1.40 (1.20, 1.62)0.680.91
      Note: Data for all-cause mortality within a 10-year follow-up period ranges from 1987 to 2019. RR (i.e., hazard ratio) was estimated from the Cox proportional hazard regression analysis.
      a Recent quitters are individuals who quit smoking <2 years ago.
      b Long-term quitters are individuals who have been quitting smoking ≥2 years.HSG, high school graduate.
      Age-specific patterns of all-cause mortality RRs by smoking intensity groups (1–5, 6–15, 16–25, and ≥26 CPD) for each race/ethnic subgroup are shown in Appendix Tables 1 and 5 (available online) and Appendix Figures 2 and 6 (available online). RRs increased with higher CPD levels for both males and females. However, some estimates, especially among individuals with ≥26 CPD, had large 95% CIs owing to small sample sizes. Among long-term quitters, RRs tended to decrease by increasing YSQ across racial/ethnic groups and gender. RRs declined to the level of never smokers for individuals with ≥15 YSQ, especially among those who were younger than age 65 years (Appendix Tables 2 and 6, available online, and Appendix Figures 3 and 7, available online).
      Age-specific all-cause mortality RRs by smoking intensity for each education group are shown in Appendix Tables 3 and 7 (available online) and Appendix Figures 4 and 8 (available online). Mortality RRs increased by increasing smoking intensity across education groups in general. RRs for long-term quitters decreased by increasing YSQ across all education groups (Appendix Tables 4 and 8, available online, and Appendix Figures 5 and 9, available online). Overall all-cause mortality RR estimates adjusted by age for each gender, race/ethnic group, and education group are also shown in Appendix Tables 9–13 (available online).

      DISCUSSION

      This study estimated all-cause mortality RRs associated with smoking in the U.S. stratified by age, gender, race/ethnicity, and educational attainment. Among current smokers or recent quitters who quit smoking <2 years ago, all-cause mortality RRs were generally higher in NHW individuals than in NHB or Hispanic individuals compared with those among never smokers. RRs also increased by the level of educational attainment, with higher RRs seen for higher-education groups. In the analyses further stratified by smoking intensity among current smokers or recent quitters, RRs generally increased in both genders with higher levels of smoking intensity across racial/ethnic and education groups. Gender differences in RRs for current smokers or recent quitters varied by age, race/ethnicity, and educational attainment. Interestingly, females tended to have higher RRs than males among individuals aged >65 years who had some college education or at least a college degree. The all-cause mortality RRs for long-term quitters were negatively associated with the length of abstinence from smoking in all racial/ethnic and education groups for both genders, with mortality rates similar to those in never smokers for individuals with 15 years or more of quitting smoking, particularly among those who were younger than age 65 years.
      Among current smokers or recent quitters, the age pattern of all-cause mortality RRs from smoking varied by gender, race/ethnicity, and educational attainment. In general, RRs increased at younger ages, peaking at age 55–64 or 65–74 years and then declining, although differences were not statistically significant for some age categories (overlapping 95% CIs). However, there were some noticeable departures from this pattern. For instance, the RRs in females with at least a college degree kept increasing by age and became highest at ages 65–74 and 75–84 years. Moreover, the RRs for long-term quitters were lower in younger ages, 30–44 and 45–54 years, than in the older age groups in both NHW males and females, but there was no clear age pattern among other racial/ethnic or education groups.
      The results are generally consistent with previous literature. A pooled cohort analysis based on follow-up time from 2000 to 2010 in 5 U.S. cohort studies showed that all-cause mortality is much higher in males than in females, but the all-cause mortality RRs between current and never smokers is similar in both genders within the same age stratum: 2.92 (95% CI=2.69, 3.18), 3.00 (95% CI=2.89, 3.13), and 2.36 (95% CI=2.24, 2.48) among males and 2.64 (95% CI=2.43, 2.86), 2.87 (95% CI=2.76, 2.99), and 2.47 (95% CI=2.37, 2.58) among females for ages 55–64, 65–74, and ≥75 years, respectively.
      HHS
      The health consequences of smoking-50 years of progress: a report of the Surgeon General.
      A recent study also based on the 1997–2018 NHIS adults aged 25–84 years who were followed up for mortality through 2019 showed that all-cause mortality RRs for current smokers versus those for never smokers were 2.91 (95% CI=2.79, 3.04), 2.21 (95% CI=2.02, 2.41), and 1.99 (95% CI=1.77, 2.24) among NHW, NHB, and Hispanic males and 3.15 (95% CI=3.03, 3.28), 2.20 (95% CI=2.02, 2.41), and 2.05 (95% CI=1.80, 2.33) among NHW, NHB, and Hispanic females, respectively.
      • Thomson B
      • Emberson J
      • Lacey B
      • et al.
      Association between smoking, Smoking Cessation, and mortality by race, ethnicity, and sex among U.S. adults.
      The RRs by age and gender for all races/ethnicity combined in this study are consistent with these studies, although this present analysis combined current smokers or recent quitters in a single group. Another study using pooled 1997–2003 NHIS adult samples aged 50–84 years linked to the National Death Index through 2006 found an all-cause mortality hazard ratio of 1.49 for NHB males relative to that for NHW males after adjusting for age at baseline, BMI, and survey year. This estimate decreased to 1.39 after additional adjustment for smoking status
      • Ho JY
      • Elo IT
      The contribution of smoking to black-white differences in U.S. Mortality.
      and decreased further to 1.15 when marital status, education, family income, and region were also adjusted. These results suggest that smoking and SES are important determinants of racial disparities in overall mortality. A study by Inoue-Choi and colleagues
      • Inoue-Choi M
      • Christensen CH
      • Rostron BL
      • et al.
      Dose-response association of low-intensity and nondaily smoking with mortality in the United States.
      showed that the RR of all-cause mortality increases by the number of CPD up to the hazard ratio of 2.94 (95% CI=2.75, 3.14) in current daily smokers who smoked >30 CPD. This is slightly lower than this study's age-adjusted hazard ratio for current smokers or recent quitters who smoked >26 CPD (hazard ratio=3.57; 95% CI=3.36, 3.78 and hazard ratio=3.99; 95% CI=3.68, 4.32 for males and females, respectively).
      Interestingly, the study findings suggest lower mortality RRs from smoking for those with lower education, who have higher smoking prevalence than those with higher education.
      • Cao P
      • Jeon J
      • Tam J
      • et al.
      Smoking disparities by level of educational attainment and birth cohort in the U.S.
      ,
      • Pampel FC
      The persistence of educational disparities in smoking.
      ,
      • Zhu BP
      • Giovino GA
      • Mowery PD
      • Eriksen MP
      The relationship between cigarette smoking and education revisited: implications for categorizing persons’ educational status.
      ,
      Current cigarette smoking among adults in the United States.
      • Huisman M
      • Kunst AE
      • Mackenbach JP
      Inequalities in the prevalence of smoking in the European Union: comparing education and income.
      • Curtin F
      • Morabia A
      • Bernstein M
      Smoking behavior in a Swiss urban population: the role of gender and education.
      • Giskes K
      • Kunst AE
      • Benach J
      • et al.
      Trends in smoking behaviour between 1985 and 2000 in nine European countries by education.
      • Tomioka K
      • Kurumatani N
      • Saeki K
      The association between education and smoking prevalence, independent of occupation: a nationally representative survey in Japan.
      • Escobedo LG
      • Peddicord JP
      Smoking prevalence in U.S. birth cohorts: the influence of gender and education.
      This apparently counterintuitive finding is likely due to greater competing mortality risks from nonsmoking sources among those with lower education, which affect all never, current, and former smokers.
      • Singh GK
      • Jemal A
      Socioeconomic and racial/ethnic disparities in cancer mortality, incidence, and survival in the United States, 1950–2014: over six decades of changing patterns and widening inequalities.
      • Businelle MS
      • Kendzor DE
      • Reitzel LR
      • et al.
      Mechanisms linking socioeconomic status to smoking cessation: a structural equation modeling approach.
      • Lewer D
      • McKee M
      • Gasparrini A
      • Reeves A
      • de Oliveira C
      Socioeconomic position and mortality risk of smoking: evidence from the English Longitudinal Study of Ageing (ELSA).
      In contrast, those with higher educational levels have fewer comorbidities or competing causes of risk, which likely then results in higher RRs when comparing people who smoke with those who never smoke. This phenomenon, in addition to differences in other relevant factors such as smoking intensity and duration,
      • Thomson B
      • Emberson J
      • Lacey B
      • et al.
      Association between smoking, Smoking Cessation, and mortality by race, ethnicity, and sex among U.S. adults.
      might also help to explain higher mortality RRs among NHW than among NHB individuals, given the higher overall mortality rates in the black than in White populations in the U.S.
      • Benjamins MR
      • Silva A
      • Saiyed NS
      • de Maio FG
      Comparison of all-cause mortality rates and inequities between black and white populations across the 30 most populous U.S. cities.
      • Haines MR
      Ethnic differences in demographic behavior in the United States has there been convergence?.
      • Harper S
      • MacLehose RF
      • Kaufman JS
      Trends in the black-white life expectancy gap among U.S. states, 1990–2009.
      • Cunningham TJ
      • Croft JB
      • Liu Y
      • Lu H
      • Eke PI
      • Giles WH
      Vital signs: racial disparities in age-specific mortality among blacks or African Americans-United States, 1999–2015.
      Smoking-related all-cause mortality RRs vary greatly by age, gender, race/ethnicity, and educational attainment. It is thus crucial to characterize the differences in RRs between key population subgroups and to consider these when assessing the impact of tobacco control interventions on smoking-related health outcomes and disparities.
      • Issabakhsh M
      • Meza R
      • Li Y
      • Yuan Z
      • Sanchez-Romero LM
      • Levy DT
      Public health impact of a U.S. menthol cigarette ban on the non-Hispanic black population: a simulation study.
      • Mendez D
      • Le TTT
      Consequences of a match made in hell: the harm caused by menthol smoking to the African American population over 1980–2018.
      HHS, NIH, National Cancer Institute
      Monograph 22: a socioecological approach to addressing tobacco-related health disparities.
      • Tam J
      • Jeon J
      • Thrasher JF
      • et al.
      Estimated prevalence of smoking and smoking-attributable mortality associated with graphic health warnings on cigarette packages in the U.S. from 2022 to 2100.
      • Apelberg BJ
      • Feirman SP
      • Salazar E
      • et al.
      Potential public health effects of reducing nicotine levels in cigarettes in the United States.
      • Le TTT
      • Mendez D
      An estimation of the harm of menthol cigarettes in the United States from 1980 to 2018.
      Not doing so can result in over/underestimation of the burden and impact of interventions in specific subgroups and mischaracterize how different interventions could affect smoking-related health disparities.
      This study used nationally representative data of the U.S. non-institutionalized population covering mortality data with a large sample from 1987 to 2019. These data provided detailed smoking information on use status, intensity, and YSQ as well as the other demographic variables, which allowed extensive analyses of all-cause mortality for people with different smoking experiences in various combinations of age, gender, and race/ethnicity or educational attainment. The estimates from this study can be readily incorporated into smoking and mortality simulation models to assess the impact of smoking tobacco control interventions on these subgroups.
      HHS
      The health consequences of smoking-50 years of progress: a report of the Surgeon General.
      ,
      • Issabakhsh M
      • Meza R
      • Li Y
      • Yuan Z
      • Sanchez-Romero LM
      • Levy DT
      Public health impact of a U.S. menthol cigarette ban on the non-Hispanic black population: a simulation study.
      • Mendez D
      • Le TTT
      Consequences of a match made in hell: the harm caused by menthol smoking to the African American population over 1980–2018.
      HHS, NIH, National Cancer Institute
      Monograph 22: a socioecological approach to addressing tobacco-related health disparities.
      The large differences in mortality patterns between subgroups found in this study suggest that group-specific estimates are essential for models to properly capture smoking risks and the impact of interventions in different sociodemographic groups.
      Previously, all-cause mortality RRs for current (by smoking intensity) and former smokers versus for never smokers by age and gender have been used as input parameters for the development of the Cancer Intervention and Surveillance Modeling Network-Lung Working Group Smoking History Generator (SHG) for the U.S. population. The SHG is a microsimulation model that generates detailed smoking histories by gender and birth cohort for the U.S. population, which are inputs for multiple simulation models.
      • Tam J
      • Levy DT
      • Jeon J
      • et al.
      Projecting the effects of tobacco control policies in the USA through microsimulation: a study protocol.
      ,
      • Jeon J
      • Holford TR
      • Levy DT
      • et al.
      Smoking and lung cancer mortality in the United States from 2015 to 2065: a comparative modeling approach.
      ,
      • Jeon J
      • Meza R
      • Krapcho M
      • Clarke LD
      • Byrne J
      • Chapter Levy DT.
      5: actual and counterfactual smoking prevalence rates in the U.S. population via microsimulation.
      • Holford TR
      • Meza R
      • Warner KE
      • et al.
      Tobacco control and the reduction in smoking-related premature deaths in the United States, 1964–2012.
      • Meza R
      • Jeon J
      • Toumazis I
      • et al.
      Evaluation of the benefits and harms of lung cancer screening with low-dose computed tomography: modeling study for the U.S. Preventive Services Task Force.
      • Cao P
      • Jeon J
      • Levy DT
      • et al.
      Potential impact of cessation interventions at the point of lung cancer screening on lung cancer and overall mortality in the United States.
      • Meza R
      • Cao P
      • Jeon J
      • et al.
      Impact of joint lung cancer screening and cessation interventions under the new recommendations of the U.S. Preventive Services Task Force.
      The new estimates in this study will facilitate the extension of the SHG and other models to simulate racial/ethnic- and educational attainment level‒specific individual smoking histories. Combined with the smoking parameters provided by Meza et al.
      • Meza R
      • Cao P
      • Jeon J
      • et al.
      Patterns of birth cohort-specific smoking histories by race and ethnicity in the U.S.
      and Cao and colleagues,
      • Cao P
      • Jeon J
      • Tam J
      • et al.
      Smoking disparities by level of educational attainment and birth cohort in the U.S.
      these all-cause mortality RRs by race/ethnicity and education will offer relevant data to inform modeling, evaluation, and surveillance tools, which can evaluate the impact of specific tobacco control interventions on smoking-related health disparities.
      HHS, NIH, National Cancer Institute
      Monograph 22: a socioecological approach to addressing tobacco-related health disparities.
      ,
      1998 Surgeon General’s report-tobacco use among U.S. Racial/ethnic minority groups | Smoking & tobacco use.
      ,
      FDA on track to take actions to address tobacco-related health disparities.

      Limitations

      This study has some limitations. Although data were combined from multiple NHIS surveys, sample sizes in some subgroups were relatively small, resulting in wide CIs. This was particularly true when estimating RRs by smoking intensity among current smokers or recent quitters and by the length of abstinence among long-term quitters. The mortality analysis did not account for underlying tobacco-related and nontobacco-related comorbidities and other health risks, which may partially explain the lower mortality RRs from smoking among lower-education or NHB individuals than among higher-education or NHW individuals. Plans to extend this analysis using the Tobacco Longitudinal Mortality Study (TLMS) are underway; this links the 1993–2019 Tobacco Use Supplement to the Current Population Survey and the National Death Index and has considerably larger sample sizes. Using the TLMS data, additional covariates can be accounted for when investigating differences in smoking-related mortality by race/ethnicity or education. Cox proportional hazard regression analyses did not adjust for any other relevant smoking determinants, such as income, U.S. region, BMI, health insurance coverage, and education (for the race/ethnicity analysis) and race/ethnicity (for the education analysis).
      • Ho JY
      • Elo IT
      The contribution of smoking to black-white differences in U.S. Mortality.
      • Lariscy JT
      • Hummer RA
      • Hayward MD
      Hispanic older adult mortality in the United States: new estimates and an assessment of factors shaping the Hispanic paradox.
      • Lariscy JT
      • Hummer RA
      • Rogers RG
      Cigarette smoking and all-cause and cause-specific adult mortality in the United States.
      • Fenelon A
      • Preston SH
      Estimating smoking-attributable mortality in the United States.
      ,
      • Inoue-Choi M
      • Christensen CH
      • Rostron BL
      • et al.
      Dose-response association of low-intensity and nondaily smoking with mortality in the United States.
      ,
      • Christensen CH
      • Rostron B
      • Cosgrove C
      • et al.
      Association of cigarette, cigar, and pipe use with mortality risk in the U.S. population.
      This is because simulation models usually do not include all of these covariates, so unadjusted estimates for specific subgroups are required to inform such models.
      HHS
      The health consequences of smoking-50 years of progress: a report of the Surgeon General.
      ,
      • Jeon J
      • Holford TR
      • Levy DT
      • et al.
      Smoking and lung cancer mortality in the United States from 2015 to 2065: a comparative modeling approach.
      ,
      • Mendez D
      • Le TTT
      Consequences of a match made in hell: the harm caused by menthol smoking to the African American population over 1980–2018.
      ,
      HHS, NIH, National Cancer Institute
      Monograph 22: a socioecological approach to addressing tobacco-related health disparities.
      ,
      • Apelberg BJ
      • Feirman SP
      • Salazar E
      • et al.
      Potential public health effects of reducing nicotine levels in cigarettes in the United States.
      • Le TTT
      • Mendez D
      An estimation of the harm of menthol cigarettes in the United States from 1980 to 2018.
      • Jeon J
      • Meza R
      • Krapcho M
      • Clarke LD
      • Byrne J
      • Chapter Levy DT.
      5: actual and counterfactual smoking prevalence rates in the U.S. population via microsimulation.
      ,
      • Levy DT
      • Tam J
      • Sanchez-Romero LM
      • et al.
      Public health implications of vaping in the USA: the smoking and vaping simulation model.
      Another limitation is that all Hispanic individuals are combined into a single group. Some studies have suggested that smoking patterns are different between Hispanic subgroups
      • Escobedo LG
      • Remington PL
      Birth cohort analysis of prevalence of cigarette smoking among Hispanics in the United States.
      ,
      • Escobedo LG
      • Remington PL
      • Anda RF
      • et al.
      Long-term secular trends in initiation of cigarette smoking among Hispanics in the United States.
      as well as between other U.S. subpopulations (e.g., Asian subpopulations and American Indian and Alaskan Native subgroups). Future analyses using larger samples (TLMS) should attempt to estimate RRs for these subpopulations.

      CONCLUSIONS

      All-cause mortality RRs associated with smoking vary by age, gender, race/ethnicity, and educational attainment. RRs are lower in Hispanic and NHB individuals than in NHW individuals and increase by level of educational attainment. The resulting estimates can be incorporated into simulation models for smoking and related health outcomes to evaluate the potential differential impact of tobacco control policies on different sociodemographic groups and on smoking-related health disparities.

      ACKNOWLEDGMENTS

      The study sponsor had no role in the study design; collection, analysis, and interpretation of data; writing of the report; or the decision to submit the report for publication.
      This project was funded through National Cancer Institute (NCI) Grants U01CA199284 and U01CA253858 and the Intramural Research Program of the NCI. The authors also acknowledge support from NCI Grant U54CA229974.
      No financial disclosures were reported by the authors of this paper.

      CRediT authorship contribution statement

      Jihyoun Jeon: Conceptualization, Funding acquisition, Methodology, Supervision, Visualization, Writing – original draft. Maki Inoue-Choi: Conceptualization, Methodology, Writing – review & editing. Yoonseo Mok: Data curation, Formal analysis, Software. Timothy S. McNeel: Data curation, Formal analysis, Software. Jamie Tam: Writing – review & editing. Neal D. Freedman: Conceptualization, Methodology, Writing – review & editing. Rafael Meza: Conceptualization, Funding acquisition, Methodology, Supervision, Writing – review & editing.

      SUPPLEMENT NOTE

      This article is part of a supplement entitled Patterns of Birth Cohort-Specific Smoking Histories by Sociodemographic Group in the U.S., which is sponsored by the National Cancer Institute (NCI) Grants U01CA199284 and U01CA253858. Authors also acknowledge support from NCI grant U54CA229974. The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of NCI.

      Appendix. SUPPLEMENTAL MATERIAL

      REFERENCES

        • HHS
        The health consequences of smoking-50 years of progress: a report of the Surgeon General.
        U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta, GA2014 (PublishedAccessed December 25, 2021)
        • U.S. National Cancer Institute
        A socioecological approach to addressing tobacco-related health disparities. National Cancer Institute Tobacco Control Monograph 22.
        HHS, NIH, National Cancer Institute, Bethesda, MD2017 (Accessed January 18, 2023.)
        • Jamal A
        • Homa DM
        • O'Connor E
        • et al.
        Current cigarette smoking among adults - United States, 2005-2014.
        MMWR Morb Mortal Wkly Rep. 2015; 64: 1233-1240https://doi.org/10.15585/mmwr.mm6444a2
        • Kanjilal S
        • Gregg EW
        • Cheng YJ
        • et al.
        Socioeconomic status and trends in disparities in 4 major risk factors for cardiovascular disease among U.S. adults, 1971–2002.
        Arch Intern Med. 2006; 166: 2348-2355https://doi.org/10.1001/archinte.166.21.2348
        • Meza R
        • Cao P
        • Jeon J
        • et al.
        Patterns of birth cohort-specific smoking histories by race and ethnicity in the U.S.
        Am J Prev Med. 2023; 64: S11-S21https://doi.org/10.1016/j.amepre.2022.06.022
        • Cao P
        • Jeon J
        • Tam J
        • et al.
        Smoking disparities by level of educational attainment and birth cohort in the U.S.
        Am J Prev Med. 2023; 64: S22-S31https://doi.org/10.1016/j.amepre.2022.06.021
        • Jeon J
        • Cao P
        • Fleischer NL
        • et al.
        Birth cohort-specific smoking patterns by family income in the U.S.
        Am J Prev Med. 2023; 64: S32-S41https://doi.org/10.1016/j.amepre.2022.07.019
        • Ho JY
        • Elo IT
        The contribution of smoking to black-white differences in U.S. Mortality.
        Demography. 2013; 50: 545-568https://doi.org/10.1007/s13524-012-0159-z
        • Lariscy JT
        • Hummer RA
        • Hayward MD
        Hispanic older adult mortality in the United States: new estimates and an assessment of factors shaping the Hispanic paradox.
        Demography. 2015; 52: 1-14https://doi.org/10.1007/s13524-014-0357-y
        • Lariscy JT
        • Hummer RA
        • Rogers RG
        Cigarette smoking and all-cause and cause-specific adult mortality in the United States.
        Demography. 2018; 55: 1855-1885https://doi.org/10.1007/s13524-018-0707-2
        • Fenelon A
        • Preston SH
        Estimating smoking-attributable mortality in the United States.
        Demography. 2012; 49: 797-818https://doi.org/10.1007/s13524-012-0108-x
        • Holford TR
        • Levy DT
        • Meza R
        Comparison of smoking history patterns among African American and white cohorts in the United States born 1890 to 1990.
        Nicotine Tob Res. 2016; 18: S16-S29https://doi.org/10.1093/ntr/ntv274
        • Ho JY
        • Fenelon A
        The contribution of smoking to educational gradients in U.S. Life expectancy.
        J Health Soc Behav. 2015; 56: 307-322https://doi.org/10.1177/0022146515592731
      1. Tobacco-related mortality. Centers for Disease Control and Prevention.https://www.cdc.gov/tobacco/data_statistics/fact_sheets/health_effects/tobacco_related_mortality/index.htm. Accessed January 18, 2023.

        • Pampel FC
        The persistence of educational disparities in smoking.
        Soc Probl. 2009; 56: 526-542https://doi.org/10.1525/sp.2009.56.3.526
        • Maralani V
        Understanding the links between education and smoking.
        Soc Sci Res. 2014; 48: 20-34https://doi.org/10.1016/j.ssresearch.2014.05.007
        • Zhu BP
        • Giovino GA
        • Mowery PD
        • Eriksen MP
        The relationship between cigarette smoking and education revisited: implications for categorizing persons’ educational status.
        Am J Public Health. 1996; 86: 1582-1589https://doi.org/10.2105/AJPH.86.11.1582
        • Rostron BL
        • Lynn BCD
        • Chang CM
        • Ren C
        • Salazar E
        • Ambrose BK
        The contribution of smoking-attributable mortality to differences in mortality and life expectancy among U.S. African-American and white adults, 2000–2019.
        Demogr Res. 2022; 46: 905-918https://doi.org/10.4054/DemRes.2022.46.31
        • Inoue-Choi M
        • Christensen CH
        • Rostron BL
        • et al.
        Dose-response association of low-intensity and nondaily smoking with mortality in the United States.
        JAMA Netw Open. 2020; 3e206436https://doi.org/10.1001/jamanetworkopen.2020.6436
        • Moolgavkar SH
        • Holford TR
        • Levy DT
        • et al.
        Impact of reduced tobacco smoking on lung cancer mortality in the United States during 1975–2000.
        J Natl Cancer Inst. 2012; 104: 541-548https://doi.org/10.1093/jnci/djs136
        • Tam J
        • Levy DT
        • Jeon J
        • et al.
        Projecting the effects of tobacco control policies in the USA through microsimulation: a study protocol.
        BMJ Open. 2018; 8e019169https://doi.org/10.1136/bmjopen-2017-019169
        • Jeon J
        • Holford TR
        • Levy DT
        • et al.
        Smoking and lung cancer mortality in the United States from 2015 to 2065: a comparative modeling approach.
        Ann Intern Med. 2018; 169: 684-693https://doi.org/10.7326/M18-1250
        • Rosenberg MA
        • Feuer EJ
        • Yu B
        • et al.
        Chapter 3: cohort life tables by smoking status, removing lung cancer as a cause of death.
        Risk Anal. 2012; 32: S25-S38https://doi.org/10.1111/j.1539-6924.2011.01662.x
        • Thun MJ
        • Carter BD
        • Feskanich D
        • et al.
        50-year trends in smoking-related mortality in the United States.
        N Engl J Med. 2013; 368: 351-364https://doi.org/10.1056/NEJMsa1211127
      2. NCHS data linkage. NCHS data linked to NDI mortality files. Centers for Disease Control and Prevention.https://www.cdc.gov/nchs/data-linkage/mortality.htm?CDC_AA_refVal=https%3A%2F%2Fwww.cdc.gov%2Fnchs%2Fdata_access%2Fdata_linkage%2Fmortality.htm. Accessed January 18, 2023.

      3. LA Blewett, JA Rivera Drew, ML King, KCW Williams, N Del Ponte, P Convey, IPUMS Health Surveys: National Health Interview Survey, version 7.2. Minneapolis, MN: IPUMS. https://doi.org/10.18128/D070.V7.2. Published 2022.

        • National Center for Health Statistics
        • Centers for Disease Control and Prevention
        The linkage of National Center for Health Statistics survey data to the National Death Index — 2019 linked mortality file (LMF): linkage methodology and analytic considerations.
        National Center for Health StatisticsCenters for Disease Control and Prevention, Atlanta, GA2022 (Published June 2, Accessed January 18, 2023.)
        • Herd N
        • Borland R
        • Hyland A.
        Predictors of smoking relapse by duration of abstinence: findings from the International Tobacco Control (ITC) Four Country Survey.
        Addiction. 2009; 104: 2088-2099https://doi.org/10.1111/j.1360-0443.2009.02732.x
        • Hughes JR
        • Peters EN
        • Naud S.
        Relapse to smoking after 1 year of abstinence: a meta-analysis.
        Addict Behav. 2008; 33: 1516-1520https://doi.org/10.1016/j.addbeh.2008.05.012
        • Krall EA
        • Garvey AJ
        • Garcia RI.
        Smoking relapse after 2 years of abstinence: findings from the VA Normative Aging Study.
        Nicotine Tob Res. 2002; 4: 95-100https://doi.org/10.1080/14622200110098428
        • Kalkhoran S
        • Kruse GR
        • Chang Y
        • Rigotti NA.
        Smoking Cessation efforts by U.S. adult smokers with medical comorbidities.
        Am J Med. 2018; 131: 318.e1-318.e8https://doi.org/10.1016/j.amjmed.2017.09.025
        • Fry JS
        • Lee PN
        • Forey BA
        • Coombs KJ.
        How rapidly does the excess risk of lung cancer decline following quitting smoking? A quantitative review using the negative exponential model.
        Regul Toxicol Pharmacol. 2013; 67: 13-26https://doi.org/10.1016/j.yrtph.2013.06.001
        • Knoke JD
        • Burns DM
        • Thun MJ.
        The change in excess risk of lung cancer attributable to smoking following smoking cessation: an examination of different analytic approaches using CPS-I data.
        Cancer Causes Control. 2008; 19: 207-219https://doi.org/10.1007/s10552-007-9086-5
        • Thomson B
        • Emberson J
        • Lacey B
        • et al.
        Association between smoking, Smoking Cessation, and mortality by race, ethnicity, and sex among U.S. adults.
        JAMA Netw Open. 2022; 5e2231480https://doi.org/10.1001/jamanetworkopen.2022.31480
      4. Current cigarette smoking among adults in the United States.
        Centers for Disease Control and Prevention, 2022 (Accessed January 18, 2023.)
        • Huisman M
        • Kunst AE
        • Mackenbach JP
        Inequalities in the prevalence of smoking in the European Union: comparing education and income.
        Prev Med (Baltimore). 2005; 40: 756-764https://doi.org/10.1016/j.ypmed.2004.09.022
        • Curtin F
        • Morabia A
        • Bernstein M
        Smoking behavior in a Swiss urban population: the role of gender and education.
        Prev Med (Baltimore). 1997; 26: 658-663https://doi.org/10.1006/pmed.1997.0187
        • Giskes K
        • Kunst AE
        • Benach J
        • et al.
        Trends in smoking behaviour between 1985 and 2000 in nine European countries by education.
        J Epidemiol Community Health. 2005; 59: 395-401https://doi.org/10.1136/jech.2004.025684
        • Tomioka K
        • Kurumatani N
        • Saeki K
        The association between education and smoking prevalence, independent of occupation: a nationally representative survey in Japan.
        J Epidemiol. 2020; 30: 136-142https://doi.org/10.2188/jea.JE20180195
        • Escobedo LG
        • Peddicord JP
        Smoking prevalence in U.S. birth cohorts: the influence of gender and education.
        Am J Public Health. 1996; 86: 231-236https://doi.org/10.2105/AJPH.86.2.231
        • Singh GK
        • Jemal A
        Socioeconomic and racial/ethnic disparities in cancer mortality, incidence, and survival in the United States, 1950–2014: over six decades of changing patterns and widening inequalities.
        J Environ Public Health. 2017; 20172819372https://doi.org/10.1155/2017/2819372
        • Businelle MS
        • Kendzor DE
        • Reitzel LR
        • et al.
        Mechanisms linking socioeconomic status to smoking cessation: a structural equation modeling approach.
        Health Psychol. 2010; 29: 262-273https://doi.org/10.1037/a0019285
        • Lewer D
        • McKee M
        • Gasparrini A
        • Reeves A
        • de Oliveira C
        Socioeconomic position and mortality risk of smoking: evidence from the English Longitudinal Study of Ageing (ELSA).
        Eur J Public Health. 2017; 27: 1068-1073https://doi.org/10.1093/eurpub/ckx059
        • Benjamins MR
        • Silva A
        • Saiyed NS
        • de Maio FG
        Comparison of all-cause mortality rates and inequities between black and white populations across the 30 most populous U.S. cities.
        JAMA Netw Open. 2021; 4e2032086https://doi.org/10.1001/jamanetworkopen.2020.32086
        • Haines MR
        Ethnic differences in demographic behavior in the United States has there been convergence?.
        Hist Methods. 2003; 36: 157-195https://doi.org/10.1080/01615440309604818
        • Harper S
        • MacLehose RF
        • Kaufman JS
        Trends in the black-white life expectancy gap among U.S. states, 1990–2009.
        Health Aff (Millwood). 2014; 33: 1375-1382https://doi.org/10.1377/hlthaff.2013.1273
        • Cunningham TJ
        • Croft JB
        • Liu Y
        • Lu H
        • Eke PI
        • Giles WH
        Vital signs: racial disparities in age-specific mortality among blacks or African Americans-United States, 1999–2015.
        MMWR Morb Mortal Wkly Rep. 2017; 66: 444-456https://doi.org/10.15585/mmwr.mm6617e1
        • Issabakhsh M
        • Meza R
        • Li Y
        • Yuan Z
        • Sanchez-Romero LM
        • Levy DT
        Public health impact of a U.S. menthol cigarette ban on the non-Hispanic black population: a simulation study.
        Tob Control. 2022; (In press. Online June 14)https://doi.org/10.1136/tobaccocontrol-2022-057298
        • Mendez D
        • Le TTT
        Consequences of a match made in hell: the harm caused by menthol smoking to the African American population over 1980–2018.
        Tob Control. 2021; (In press. Online September 16)https://doi.org/10.1136/tobaccocontrol-2021-056748
        • HHS, NIH, National Cancer Institute
        Monograph 22: a socioecological approach to addressing tobacco-related health disparities.
        HHS, NIH, National Cancer Institute, Bethesda, MD2017 (Accessed January 18, 2023.)
        • Tam J
        • Jeon J
        • Thrasher JF
        • et al.
        Estimated prevalence of smoking and smoking-attributable mortality associated with graphic health warnings on cigarette packages in the U.S. from 2022 to 2100.
        JAMA Health Forum. 2021; 2e212852https://doi.org/10.1001/jamahealthforum.2021.2852
        • Apelberg BJ
        • Feirman SP
        • Salazar E
        • et al.
        Potential public health effects of reducing nicotine levels in cigarettes in the United States.
        N Engl J Med. 2018; 378: 1725-1733https://doi.org/10.1056/NEJMsr1714617
        • Le TTT
        • Mendez D
        An estimation of the harm of menthol cigarettes in the United States from 1980 to 2018.
        Tob Control. 2021; 31: 564-568https://doi.org/10.1136/tobaccocontrol-2020-056256
        • Jeon J
        • Meza R
        • Krapcho M
        • Clarke LD
        • Byrne J
        • Chapter Levy DT.
        5: actual and counterfactual smoking prevalence rates in the U.S. population via microsimulation.
        Risk Anal. 2012; 32: S51-S68https://doi.org/10.1111/j.1539-6924.2011.01775.x
        • Holford TR
        • Meza R
        • Warner KE
        • et al.
        Tobacco control and the reduction in smoking-related premature deaths in the United States, 1964–2012.
        JAMA. 2014; 311: 164-171https://doi.org/10.1001/jama.2013.285112
        • Meza R
        • Jeon J
        • Toumazis I
        • et al.
        Evaluation of the benefits and harms of lung cancer screening with low-dose computed tomography: modeling study for the U.S. Preventive Services Task Force.
        JAMA. 2021; 325: 988-997https://doi.org/10.1001/jama.2021.1077
        • Cao P
        • Jeon J
        • Levy DT
        • et al.
        Potential impact of cessation interventions at the point of lung cancer screening on lung cancer and overall mortality in the United States.
        J Thorac Oncol. 2020; 15: 1160-1169https://doi.org/10.1016/j.jtho.2020.02.008
        • Meza R
        • Cao P
        • Jeon J
        • et al.
        Impact of joint lung cancer screening and cessation interventions under the new recommendations of the U.S. Preventive Services Task Force.
        J Thorac Oncol. 2022; 17: 160-166https://doi.org/10.1016/j.jtho.2021.09.011
      5. 1998 Surgeon General’s report-tobacco use among U.S. Racial/ethnic minority groups | Smoking & tobacco use.
        Centers for Disease Control and Prevention, 2015 (Accessed January 18, 2023.)
      6. FDA on track to take actions to address tobacco-related health disparities.
        Food and Drug Administration, 2022 (Accessed January 18, 2023.)
        • Christensen CH
        • Rostron B
        • Cosgrove C
        • et al.
        Association of cigarette, cigar, and pipe use with mortality risk in the U.S. population.
        JAMA Intern Med. 2018; 178: 469-476https://doi.org/10.1001/jamainternmed.2017.8625
        • Levy DT
        • Tam J
        • Sanchez-Romero LM
        • et al.
        Public health implications of vaping in the USA: the smoking and vaping simulation model.
        Popul Health Metr. 2021; 19: 19https://doi.org/10.1186/s12963-021-00250-7
        • Escobedo LG
        • Remington PL
        Birth cohort analysis of prevalence of cigarette smoking among Hispanics in the United States.
        JAMA. 1989; 261: 66-69https://doi.org/10.1001/jama.1989.03420010076036
        • Escobedo LG
        • Remington PL
        • Anda RF
        • et al.
        Long-term secular trends in initiation of cigarette smoking among Hispanics in the United States.
        Public Health Rep. 1989; 104 (Accessed January 18, 2023): 583-587https://doi.org/10.15585/mmwr.mm6444a2