Global burden of lung cancer attributable to secondhand smoke (1990–2021) and predictive trends to 2035 Authors: Xiao Wang, Yezu Liu, Liaoliao Bai, Hongqing Chen, Ailin Zhu, Haizhou Dang Published: 2026 ABSTRACT Secondhand smoke (SHS) is closely related to lung cancer, however, the global burden of lung cancer caused by SHS remains unclear. This study used 2021 Global Burden of Disease data (GBD) to comprehensively assess the lung cancer burden associated with SHS. The number of deaths and disability-adjusted life years (DALYs) attributable to SHS was found to be 0.098 million and 2.4 million, respectively in 2021. Nevertheless, the burden of lung cancer attributable to SHS decreased from 1990 to 2019. At the sociodemographic index (SDI) district level, the middle-high SDI region had the greatest number of lung cancer deaths and DALYs attributable to SHS. Conversely, in high-SDI regions, the age-standardized mortality rateand age-standardized DALYs (age-standardized daly rate) rate of lung cancer attributable to SHS exhibit the most rapid declines. Moreover, the burden of lung cancer was mainly distributed in East Asia. The burden of lung cancer attributable to SHS is heavier for men than for women. Conversely, significant absolute and relative inequalities in lung cancer burden attributable to SHS, as measured by the SDI, were observed, with a marked decrease in these disparities over time. Globally, the number of deaths and DALYs is projected to increase, while the age-standardized rates (ASR) for both mortality and DALYs are expected to decline annually through 2035. From 1990 to 2021, an upward trend in the absolute number of lung cancer deaths and DALYs attributable to SHS was accompanied by a concurrent downward trend in their ASR. Inequalities associated with the SDI narrowed across countries over time. While this study has limitations inherent to the GBD database, these findings nonetheless underscore the importance of policies that protect nonsmokers from SHS exposure. 1. Introduction Cancer is the second leading cause of death worldwide, imposing enormous physical, mental, and economic strain on societies globally. As lung cancer cases and deaths continue to rise, it has become the most commonly diagnosed cancer and the leading cause of cancer-related mortality[1] – a trend that warrants growing attention. Based on the 2021 Global Burden of Disease (GBD) statistics, lung cancer ranks as the 17th leading cause of disability-adjusted life years (DALYs) overall. Among age groups, it is the fifth leading cause of DALYs in individuals aged 50 to 74 and the seventh leading cause among those aged 75 and older.[2] Smoking remains the biggest risk factor for lung cancer, exceeding all other factors causing lung cancer.[3] Tobacco and tobacco smoke contain a complex mixture of over 9500 compounds. The International Agency for Research on Cancer has identified at least 80 carcinogens in tobacco smoke.[4] Exposure to secondhand smoke (SHS) has been classified by the International Agency for Research on Cancer as a Group 1 carcinogen (known human carcinogen). In numerous epidemiological studies, passive smoking is scientifically undisputed as one of the causes of lung cancer.[5] An estimated 1 million people die globally each year from passive smoking.[6] SHS refers to passive smoking, which means nonsmokers are exposed to mainstream smoke and sidestream smoke due to others’ smoking,[7]Containing carcinogens: benzene, benzo-a-pyrene, and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone[3]; the urine from nonsmokers exposed to SHS contains measurable carcinogenic metabolites.[8] Related studies have indicated a correlation between SHS exposure and increased morbidity and mortality,[9] and in the United States, annual productivity losses due to SHS exposure are estimated at $6.6 billion.[10] A comprehensive understanding of the epidemiological features of secondhand smoke–attributable lung cancer at the global, regional, and national levels remains limited.[2,11] Based on data from the Global Burden of Disease 2021 study, this study accounted for regional, sex, and age differences, projected the mortality and disability burden of lung cancer attributable to SHS to 2035, and further quantified its association with SHS. Specifically, this study included descriptive epidemiological analyses of lung cancer due to SHS at global, regional, and national levels; trend analyses at global, local, and multidimensional scales; cross-country inequality analysis according to the standard health equity analysis method recommended by the World Health Organization; and prediction analysis of the global lung cancer burden attributable to SHS to 2035. This study provides relevant references for the development of prevention and control strategies for secondhand smoke-related lung cancer. 2. Materials and methods 2.1. Study data Our research utilized the data from the 2021 GBD study. Two primary indicators were examined: mortality rates and DALYs resulting from lung cancer attributable to SHS. Our analysis covered 204 countries and regions, which were divided into 21 GBD regions based on geographical proximity and further categorized into 5 groups based on the socio-demographic index (SDI). In the GBD 2021 database, lung cancer is a tumor formed in the trachea, bronchus, or lung. ICD-10 codes for lung cancer include C33 to C34.9, D02.1 to D02.3, D14.2 to D14.3, and D38.1. SHS was defined exposure as current exposure to secondhand tobacco smoke at home or at work. And only consider nonsmokers to be exposed to SHS. nonsmokers are defined as all persons who are not daily smokers. Ex-smokers and occasional smokers are considered nonsmokers in this analysis.[11] Estimates of mortality and DALYs from lung cancer attributable to SHS were calculated by multiplying the population attributable fraction – stratified by location, year, age, and sex – by the corresponding lung cancer mortality or DALYs. We used DALYs as the primary disease burden measure, which combines years of life lost from premature mortality with years lived with disability. The calculation of years of life lost involved multiplying age-specific lung cancer deaths by the remaining life expectancy for that age group, while years lived with disability were obtained as the product of lung cancer prevalence and severity-specific disability weights. The SDI, a standardized indicator in GBD analyses, was used to assess the association between disease burden and socioeconomic development. This composite index ranges from 0 to 1 and captures 3 key dimensions: per capita income, average educational attainment among individuals aged 15 years and older, and total fertility rate among those under 25 years. Better socioeconomic conditions and health outcomes are associated with higher SDI values. According to the SDI, regions fall into 5 tiers: low (<0.46), low-middle (0.46–0.64), middle (0.65–0.74), high-middle (0.75–0.85), and high (>0.85). 2.2. Statistical analysis The analysis commenced with an exploration of the dataset structure, followed by the computation of counts and rates for SHS‑related lung cancer mortality and DALYs at global, regional, and national scales. Next, we evaluated regional trends in these indicators over the period 1990 to 2021. We utilized age-standardized rates (ASRs) to facilitate comparisons of mortality and DALY rates across nations, accounting for differences in age composition and population characteristics. An increasing ASR trend was inferred when both the estimated annual percentage change (EAPC) and the lower 95% CI bound exceeded zero; a decreasing trend was inferred when both the EAPC and the upper 95% CI bound were below zero. In all other cases, the ASR was considered stable throughout the study period. The EAPC was computed as 100×(eβ−1), together with a 95% confidence interval (95% CI). Temporal changes were evaluated using a linear regression on the natural logarithms of the rates: y=α+βx+ε, where x is the year and y=ln(rate).[12,13] The correlation between ASR and the SDI was evaluated using Spearman’s rank correlation test.[13,14] We explored the trend of lung cancer attributable to SHS burden associated with age, period, and birth cohort effects. Widely used across demography, sociology, and epidemiology, the age-period-cohort (APC) model serves as a statistical tool to distinguishes 3 dimensions: age effects, reflecting risk variations across age groups; period effects, capturing time‑related mortality changes that impact all ages simultaneously; and cohort effects, representing differences among birth cohorts.[15] The APC analysis is inherently constrained by the identification problem arising from the perfect multicollinearity among age, period, and cohort. To resolve this, we applied an APC model with the intrinsic estimator, a method based on principal component regression. This approach effectively decomposes the variance in the data into the distinct effects of age, period, and cohort, providing robust and estimable effects for each dimension.[16] The age-period-cohort model with intrinsic estimator method based on the Poisson distribution could be expressed as the following: ln⁡(Mijk)==ln⁡(DijkPijk)=μ+αi+βj+γk+εijk. The model defines Mijk as the incidence or mortality of SHS-attributable lung cancer for age group i, time period j, and birth cohort k. Correspondingly, Dijk and Pijk indicate the number of cases (or deaths) and the population at risk, respectively. The intercept is µ, and αi, βj, and γk are the coefficients for age, period, and cohort. The random error εijk is assumed to follow a standard normal distribution.[17] Two standard indicators of absolute and relative gradient inequality – the slope index of inequality (SII) and the concentration index (CI) – were used to quantify cross-country disparities in the burden of SHS-attributable lung cancer.[18,19] The SII was estimated by regressing national all-age DALY rates on a relative position scale linked to sociodemographic development. The CI was computed by integrating the area under the Lorenz concentration curve, which was constructed using the cumulative distribution of DALYs and the cumulative population ranked by SDI.[20] The monitoring of health inequities reveals disparities in the burden of lung cancer attributable to secondhand smoke across countries in different SDI regions, offering a basis for policy improvement. To project the trends of secondhand smoke‑related lung cancer through 2035, the Bayesian age‑period‑cohort (BAPC) model was used in this study.[21] Incorporating age, period, and cohort effects, this model leverages historical data and probability distributions to forecast future SHS‑attributable lung cancer patterns. All data processing, statistical analyses, and visualizations were conducted using R software (version 4.4.1). 2.3. Ethics This study used anonymized, publicly accessible data from the 2021 GBD database; therefore, ethical approval was deemed unnecessary, and informed consent was waived. 3. Results 3.1. The spatiotemporal patterns of the burden of lung cancer caused by SHS Globally in 2021, lung cancer attributable to SHS resulted in 0.098 million deaths and 2.4 million DALYs, with a male-to-female ratio of approximately 1.4 for both measures. A substantial upward trajectory was observed in the total counts of deaths and DALYs from SHS-attributable lung cancer over the period 1990 to 2021, whereas the age-standardized mortality rate (ASMR) and age-standardized DALY rate (ASDR) showed a declining trend over the same period. The EAPC of ASMR and ASDR was significantly higher in males than females (Table 1 and Fig. S1A and B). Table 1Global and regional burden of lung cancer attributable to secondhand smoke in 1990 and 2021, and EAPC of both ASMR and ASDR from 1990 to 2021.Characteristics19902021EAPC (1990–2021)Deaths casesASMRDALYsASDRDeaths casesASMRDALYsASDRASMRASDR(×103) (95% UI)(per 100, 000) (95% UI)(×104) (95% UI)(per 100,000) (95% UI)(×103) (95% UI)(per 100,000) (95% UI)(×104) (95% UI)(per 100,000) (95% UI)(95% UI)(95% UI)Overall57.62 (7.09–108.32)2.62 (0.32–4.94)159.89 (19.72–299.55)69.55 (8.57–130.42)97.91 (12.22–187.95)2.06 (0.26–3.96)235.59 (29.62–448.85)48.76 (6.13–92.98)−0.88 (−0.94–−0.82)−1.25 (−1.31–−1.19)Sex Male36.76 (4.38–70.65)3.67 (0.44–7.05)102.09 (12.22–196.04)93.96 (11.22–180.47)56.85 (6.68–112.37)2.62 (0.31–5.19)135.96 (16.04–266.29)59.54 (7.03–116.88)−1.07 (−1.15–−1)−1.48 (−1.55–−1.4) Female20.85 (2.76–38.42)1.77 (0.23–3.27)57.8 (7.62–105.93)48.09 (6.34–88.17)41.06 (5.37–80.16)1.61 (0.21–3.14)99.63 (13.07–193.76)39.49 (5.19–76.8)−0.57 (−0.65–−0.49)−0.89 (−0.97–−0.81)SDl region High SDI18.92 (2.35–36.17)3.16 (0.39–6.03)50.77 (6.33–96.57)87.16 (10.86–165.55)16.95 (2.19–33.12)1.48 (0.19–2.88)38.03 (4.97–73.81)36.2 (4.74–70.07)−2.57 (−2.67–−2.46)−2.93 (−3.04–−2.81) High·middle SDI21.7 (2.61–41.37)3.91 (0.47–7.45)60.93 (7.36–115.91)106.3 (12.83–202.24)39.12 (4.78–75.12)3.55 (0.43–6.82)93.66 (11.53–178.47)85.48 (10.55–163.02)−0.36 (−0.46–−0.25)−0.78 (−0.89–−0.67) Middle SDI14.19 (1.78–26.81)2.55 (0.32–4.82)40.12 (5.05–75.81)64.84 (8.15–122.51)35.51 (4.49–69.3)2.46 (0.31–4.82)86.23 (11–166.85)56.39 (7.18–109.39)−0.24 (−0.3–−0.18)−0.58 (−0.63–−0.52) Low·middle SDl2.35 (0.31–4.4)0.71 (0.09–1.33)6.75 (0.87–12.63)18.38 (2.39–34.43)5.48 (0.7–10.51)0.69 (0.09–1.33)15.23 (1.93–29.2)17.93 (2.27–34.39)−0.13 (−0.17–−0.09)−0.13 (−0.17–−0.09) Low SDl0.37 (0.04–0.75)0.3 (0.03–0.61)1.08 (0.12–2.17)7.91 (0.9–15.84)0.77 (0.09–1.54)0.28 (0.03–0.56)2.23 (0.27–4.48)7.18 (0.86–14.36)−0.43 (−0.52–−0.35)−0.49 (−0.58–−0.4)GBD region East Asia21.36 (2.6–41.46)4.7 (0.57–9.12)59.84 (7.34–116)116.65 (14.25–226.18)59.2 (7.24–116.46)4.98 (0.61–9.83)138.8 (17.23–270.88)113.06 (14.03–221.23)0.14 (0.01–0.26)−0.19 (−0.29–−0.08) Southeast Asia2.12 (0.28–4.05)1.56 (0.21–2.98)5.94 (0.79–11.32)39.22 (5.22–74.83)5.33 (0.7–10.36)1.5 (0.2–2.94)14.22 (1.86–27.62)36.8 (4.83–71.58)−0.33 (−0.42–−0.24)−0.41 (−0.5–−0.32) Oceania0.02 (0–0.04)1.22 (0.14–2.59)0.05 (0.01–0.1)28.8 (3.36–61.34)0.05 (0.01–0.11)1.36 (0.17–2.96)0.14 (0.02–0.31)32.14 (3.87–70.28)0.3 (0.23–0.38)0.29 (0.21–0.37) Central Asia0.71 (0.09–1.41)2.63 (0.33–5.24)2.14 (0.27–4.26)75.97 (9.5–151.41)0.6 (0.08–1.2)1.31 (0.17–2.63)1.68 (0.22–3.35)33.96 (4.42–67.82)−1.8 (−1.95–−1.65)−2.24 (−2.36–−2.12) Central Europe3.81 (0.48–7.25)4.54 (0.57–8.64)11.06 (1.38–21.01)131.71 (16.49–250.22)3.49 (0.45–6.83)2.92 (0.37–5.7)8.62 (1.09–16.82)76.62 (9.67–149.1)−1.46 (−1.64–−1.27)−1.8 (−2.01–−1.59) Eastern Europe4.32 (0.55–8.42)2.72 (0.35–5.3)12.57 (1.59–24.48)79.08 (9.96–153.86)2.83 (0.33–5.7)1.46 (0.17–2.93)7.54 (0.9–15.16)40.27 (4.84–80.73)−2.1 (−2.26–−1.93)−2.3 (−2.47–−2.13) High·income Asia Pacific2.64 (0.33–5.19)2.39 (0.3–4.72)6.6 (0.82–12.85)57.96 (7.17–112.99)3.44 (0.41–7.21)1.24 (0.15–2.58)6.27 (0.76–13.06)26.86 (3.28–55.53)−2.41 (−2.64–−2.19)−2.74 (−2.97–−2.51) Australasia0.28 (0.03–0.62)2.21 (0.23–4.85)0.76 (0.08–1.65)60.43 (6.31–131.32)0.24 (0.03–0.56)0.85 (0.09–1.94)0.58 (0.06–1.32)21.94 (2.3–49.44)−3.04 (−3.14–−2.95)−3.2 (−3.28–−3.12) Western Europe9.03 (1.07–17.36)2.98 (0.36–5.72)24.47 (2.93–46.87)85.05 (10.2–162.54)6.23 (0.78–12.34)1.32 (0.17–2.6)15.08 (1.9–29.73)35.25 (4.45–69.38)−2.59 (−2.68–−2.49)−2.79 (−2.91–−2.67) Southern Latin America0.69 (0.08–1.44)2.7 (0.33–5.63)1.94 (0.24–4.05)75.07 (9.23–156.42)0.61 (0.07–1.31)1.28 (0.15–2.73)1.54 (0.18–3.28)33.14 (3.98–70.26)−2.31 (−2.53–−2.09)−2.57 (−2.81–−2.33) High·income North America7.02 (0.83–13.65)3.79 (0.45–7.36)18.81 (2.24–36.36)106.27 (12.67–205)4.57 (0.56–9.18)1.26 (0.15–2.54)10.74 (1.3–21.46)31.44 (3.82–62.72)−3.75 (−3.9–−3.6)−4.11 (−4.26–−3.96) Caribbean0.27 (0.03–0.55)1.93 (0.22–3.95)0.63 (0.07–1.28)43.45 (5.07–88.45)0.36 (0.04–0.76)1.19 (0.13–2.53)0.81 (0.09–1.7)27.1 (2.97–57.09)−1.56 (−1.73–−1.39)−1.59 (−1.78–−1.4) Andean Latin America0.05 (0.01–0.09)0.43 (0.05–0.83)0.13 (0.02–0.26)10.98 (1.35–21.23)0.07 (0.01–0.14)0.21 (0.02–0.42)0.17 (0.02–0.35)5.14 (0.57–10.51)−2.9 (−3.18–−2.63)−3.02 (−3.31–−2.72) Central Latin America0.34 (0.04–0.64)0.77 (0.09–1.46)0.89 (0.11–1.67)18.44 (2.25–34.77)0.46 (0.06–0.88)0.33 (0.04–0.65)1.12 (0.14–2.17)7.95 (0.97–15.37)−3 (−3.12–−2.89)−3.02 (−3.14–−2.9) Tropical Latin America0.92 (0.11–1.77)1.85 (0.22–3.58)2.54 (0.31–4.91)47.12 (5.73–91.16)1.37 (0.16–2.82)0.96 (0.11–1.99)3.44 (0.4–7.08)23.68 (2.72–48.77)−2.3 (−2.4–−2.19)−2.44 (−2.53–−2.34) North Africa and the Middle East2.15 (0.24–4.15)2.38 (0.27–4.59)6.02 (0.67–11.61)59.95 (6.69–115.6)4.53 (0.53–9.03)1.89 (0.22–3.77)12.01 (1.4–23.84)44.97 (5.26–89.44)−0.78 (−0.89–−0.66)−0.99 (−1.09–−0.89) South Asia1.48 (0.19–2.85)0.48 (0.06–0.92)4.28 (0.55–8.21)12.27 (1.57–23.57)3.87 (0.51–7.76)0.47 (0.06–0.95)10.83 (1.4–21.69)12.4 (1.61–24.85)−0.42 (−0.6–−0.24)−0.35 (−0.52–−0.17) Central Sub·Saharan Africa0.04 (0–0.08)0.29 (0.03–0.64)0.12 (0.01–0.25)8.1 (0.94–17.45)0.08 (0.01–0.18)0.26 (0.03–0.57)0.27 (0.03–0.58)7.23 (0.77–15.59)−0.44 (−0.75–−0.13)−0.42 (−0.73–−0.11) Eastern Sub·Saharan Africa0.09 (0.01–0.18)0.21 (0.03–0.42)0.27 (0.03–0.55)5.87 (0.71–11.7)0.14 (0.02–0.28)0.15 (0.02–0.29)0.43 (0.05–0.84)3.99 (0.5–7.77)−1.36 (−1.45–−1.26)−1.48 (−1.59–−1.38) Southern Sub·Saharan Africa0.22 (0.03–0.44)1.47 (0.18–2.91)0.67 (0.08–1.33)40.84 (4.85–80.8)0.31 (0.04–0.6)0.95 (0.12–1.84)0.9 (0.11–1.74)25.49 (3.24–49.49)−1.48 (−1.7–−1.27)−1.56 (−1.78–−1.35) Western Sub·Saharan Africa0.06 (0.01–0.13)0.14 (0.02–0.27)0.18 (0.02–0.35)3.43 (0.44–6.78)0.14 (0.02–0.28)0.14 (0.02–0.27)0.4 (0.05–0.79)3.35 (0.41–6.65)−0.05 (−0.13–0.03)−0.15 (−0.23–−0.06)ASDR = age·standardized DALYS rate, ASMR = age·standardized mortality rate, Cl = confidential interval, DALY = disability·adjusted life·year, EAPC = estimated annual percentage change, GBD = Global Burden of Disease, SDl = sociodemographic index, UI = uncertainty interval. Global and regional burden of lung cancer attributable to secondhand smoke in 1990 and 2021, and EAPC of both ASMR and ASDR from 1990 to 2021. ASDR = age·standardized DALYS rate, ASMR = age·standardized mortality rate, Cl = confidential interval, DALY = disability·adjusted life·year, EAPC = estimated annual percentage change, GBD = Global Burden of Disease, SDl = sociodemographic index, UI = uncertainty interval. Across SDI strata, the highest absolute burden of SHS‑attributable lung cancer in 2021 was concentrated in upper‑middle and high SDI regions, with 0.039 million deaths and 0.94 million DALYs, representing approximately 40% of the global total. Conversely, low SDI regions recorded the highest ASMR and ASDR. From 1990 to 2021, all SDI regions showed a consistent decline in both ASMR and ASDR, with the most pronounced reductions observed in high SDI regions (ASMR EAPC: −2.57, 95% CI: −2.6–−2.46; ASDR EAPC: −2.93, 95% CI: −3.04–−2.81) (Table 1; Fig. S2A and B). Across all SDI categories, males consistently exhibited higher age‑standardized rates. Notable sex‑based disparities persisted in low and low‑middle SDI regions. While these disparities narrowed over time in high and upper‑middle SDI regions, they widened in middle SDI regions; no significant temporal changes were detected in low and low‑middle SDI regions (Fig. S3A and B). By GBD region, East Asia accounted for the largest share of global lung cancer mortality and DALY burden in 2021, comprising over 58% of the worldwide total, and also maintained the highest ASMR and ASDR levels (Table 1; Fig. 1A and B). Between 1990 and 2021, Australasia, High‑income North America, and Central Latin America saw more rapid reductions in both ASMR and ASDR, with all corresponding EAPC values exceeding −3. In contrast, increasing trends were observed in East Asia and Oceania: the EAPC for ASDR was 0.14 (95% CI: 0.01–0.26) in East Asia, and in Oceania, the EAPC for ASDR was 0.3 (95% CI: 0.23–0.38) while that for ASMR was 0.29 (95% CI: 0.21–0.37) (Table 1; Fig. 1C and D). Across 22 global regions, sex-specific estimates for SHS-attributable lung cancer: (A) ASMR in 2021; (B) ASDR in 2021; (C) EAPC in ASMR from 1990 to 2021; (D) EAPC in ASDR from 1990 to 2021. Blue: males; red: females. ASDR = age-standardized DALY rate, ASMR = age-standardized mortality rate, DALYs = disability-adjusted life-years, EAPC = estimated annual percentage change, SHS = secondhand smoke. China had the highest counts of SHS‑attributable lung cancer deaths and DALYs in 2021. Regarding ASRs, Montenegro, China, and North Macedonia topped the ASMR ranking, while Turkey, Hungary, and Bulgaria led in ASDR (Fig. 2A and B). The fastest declines were seen in Mexico for ASMR (EAPC: −4.63; 95% CI: −4.83–−4.42) and in the United Kingdom for ASDR (EAPC: −4.79; 95% CI: −4.86–−4.73) (Table S1; Fig. 2C and D). For SHS-attributable lung cancer, shown across the globe are (A) ASMR and (B) ASDR for 2021, and (C) and (D) the EAPC in ASMR and ASDR over 1990–2021. ASDR = age-standardized DALY rate, ASMR = age-standardized mortality rate, DALYs = disability-adjusted life-years, EAPC = estimated annual percentage change, SHS = secondhand smoke. 3.2. Global lung cancer burden attributable to SHS by age and sex Figure S4 illustrates global age‑specific mortality rates for SHS‑attributable lung cancer in 2021. Distinct sex‑based patterns emerged: mortality rates for both sexes rose progressively until peaking in the 85 to 89 age group; thereafter, rates declined sharply among males while remaining relatively stable among females. Across all ages, male mortality consistently exceeded female mortality, with the widest sex gap observed in the 85 to 89 age bracket (Fig. S4A). The age‑specific DALY burden followed a parabolic trajectory similar to that of mortality, with rates increasing across age groups, peaking at 75 to 79 years, and then declining in both sexes. The largest sex disparity in DALYs also occurred in the 85 to 89 age group (Fig. S4B). From 1990 to 2021, global mortality rates declined for individuals under 79 years of age, with the most rapid reduction observed in the 35 to 39 age bracket, whereas rates increased among those aged over 79. Temporal trends by sex mirrored the overall pattern across all age groups (Fig. S5A). Across SDI regions, mortality rates consistently rose after a certain age threshold; notably, rates in medium‑low SDI regions remained relatively stable across all age groups (Fig. S5B). Parallel trends were evident in the estimated annual percentage changes (EAPC) for age‑specific DALY ratios (Fig. S6A and B). 3.3. Association with SDI Figure 3 presents the regional and national correlations between ASMR for SHS-attributable lung cancer and SDI. Regionally, from 1990 to 2021, ASMR and ASDR were negatively associated with SDI in most regions, except East Asia and Oceania. The global ASMR fell below the SDI-expected level, whereas East Asia, Central Europe, and High-income North America showed observed ASMR values above expectations (Fig. 3A). Nationally in 2021, ASMR was positively correlated with SDI (ρ = 0.53, P < .001); this positive association persisted across low-medium, medium, and medium-high SDI categories but attenuated in the high SDI category. ASMR and ASDR were generally lower in low SDI regions than in high SDI regions. East Asia and Central Europe had observed rates markedly above SDI-based expectations, while Central Latin America and Andean Latin America had rates well below them (Fig. 3B). Additionally, EAPC in ASMR showed a negative correlation with the 1990 ASMR (ρ = –0.34, P < .001) (Fig. S7A). A strong negative correlation between EAPC in ASMR and SDI was identified in 2021 (ρ = –0.51, P < .001), which became more pronounced at SDI levels above 0.5 – particularly in countries with SDI between 0.75 and 0.9 (Fig. S7B). Parallel patterns were found for ASDR (Fig. S8A and B; Fig. S9A and B). Shown are the associations between ASMR for SHS-attributable lung cancer and SDI at regional (A) and national (B) levels. Across global and 21 regions, panel A displays the temporal trend in ASMR by SDI from 1990 to 2021; panel B depicts the ASMR by SDI for 204 countries and territories in 2021. The black line denotes the ASMR predicted by SDI. ASMR = age-standardized mortality rate, SDI = sociodemographic index, SHS = secondhand smoke. 3.4. Age-period-cohort analysis on mortality and DALYs of lung cancer attributable to SHS With period and cohort effects controlled for, age emerged as a significant determinant of mortality and DALYs from SHS‑attributable lung cancer (Fig. 4). The relative risks followed a parabolic pattern: mortality peaked at 85 to 89 years, while DALYs peaked at 70 to 74 years (Table 2). Sex-stratified analyses (Figs. S10 and S11 ) revealed that male mortality consistently exceeded female mortality after age 50, and a similar male predominance in DALYs was observed from age 50 to 95 years (Table S2). Table 2RRs of lung cancer attributable to secondhand smoke deaths and DALYs for both sexes due to age, period, and birth cohort effects.FactorDeathsDALYsRR (95 % CI) P RR (95 % CI) P Age (yr) 25–290.053 (0.05–0.057)<.0010.097 (0.096–0.098)<.001 30–340.098 (0.094–0.102)<.0010.172 (0.171–0.173)<.001 35–390.181 (0.176–0.187)<.0010.304 (0.302–0.306)<.001 40–440.364 (0.356–0.373)<.0010.581 (0.578–0.584)<.001 45–490.653 (0.641–0.666)<.0010.979 (0.976–0.983)<.001 50–541.135 (1.118–1.152)<.0011.583 (1.579–1.588)<.001 55–591.659 (1.638–1.681)<.0012.126 (2.12–2.131)<.001 60–642.05 (2.026–2.074)<.0012.367 (2.362–2.373)<.001 65–692.325 (2.298–2.351)<.0012.362 (2.355–2.368)<.001 70–742.756 (2.723–2.789)<.0012.406 (2.398–2.414)<.001 75–793.031 (2.99–3.072)<.0012.212 (2.204–2.221)<.001 80–843.136 (3.086–3.187)<.0011.874 (1.865–1.883)<.001 85–893.202 (3.139–3.266)<.0011.588 (1.578–1.598)<.001 90–942.755 (2.68–2.831)<.0011.241 (1.23–1.252)<.001 95–992.134 (2.032–2.241)<.0010.933 (0.917–0.949)<.001Period 19920.852 (0.843–0.86)<.0010.947 (0.944–0.949)<.001 19970.898 (0.891–0.906)<.0010.96 (0.958–0.962)<.001 20020.957 (0.95–0.963)<.0010.982 (0.98–0.983)<.001 20071.021 (1.014–1.027)<.0011.002 (1.001–1.003).002 20121.103 (1.095–1.112)<.0011.033 (1.031–1.035)<.001 20171.213 (1.202–1.225)<.0011.083 (1.08–1.086)<.001Birth cohort 1897–19012.207 (1.847–2.637)<.0011.477 (1.389–1.572)<.001 1902–19062.159 (2.003–2.327)<.0011.49 (1.453–1.529)<.001 1907–19112.037 (1.946–2.133)<.0011.465 (1.443–1.487)<.001 1912–19162.034 (1.963–2.108)<.0011.526 (1.509–1.544)<.001 1917–19212.067 (2.005–2.131)<.0011.617 (1.601–1.633)<.001 1922–19261.999 (1.946–2.053)<.0011.641 (1.627–1.655)<.001 1927–19311.933 (1.887–1.98)<.0011.658 (1.645–1.671)<.001 1932–19361.791 (1.752–1.831)<.0011.595 (1.584–1.606)<.001 1937–19411.565 (1.534–1.597)<.0011.455 (1.446–1.463)<.001 1942–19461.348 (1.323–1.375)<.0011.31 (1.303–1.317)<.001 1947–19511.193 (1.171–1.216)<.0011.215 (1.21–1.221)<.001 1952–19561.06 (1.04–1.08)<.0011.127 (1.123–1.132)<.001 1957–19610.869 (0.853–0.887)<.0010.965 (0.962–0.968)<.001 1962–19660.698 (0.683–0.713)<.0010.814 (0.811–0.816)<.001 1967–19710.596 (0.582–0.61)<.0010.726 (0.724–0.729)<.001 1972–19760.465 (0.451–0.478)<.0010.592 (0.59–0.595)<.001 1977–19810.366 (0.352–0.381)<.0010.488 (0.486–0.491)<.001 1982–19860.309 (0.293–0.326)<.0010.432 (0.429–0.435)<.001 1987–19910.293 (0.272–0.316)<.0010.428 (0.424–0.433)<.001 1992–19960.238 (0.205–0.277)<.0010.363 (0.356–0.37)<.001CI = confidence interval, DALY = disability·adjusted life·year, RRs = relative risk. RRs of lung cancer attributable to secondhand smoke deaths and DALYs for both sexes due to age, period, and birth cohort effects. CI = confidence interval, DALY = disability·adjusted life·year, RRs = relative risk. The effects of age, period, and birth cohort on the relative risk of lung cancer attributable to secondhand smoke mortality and (A) and DALYs (B). DALYs = disability-adjusted life-years. With age and cohort effects controlled for, a significant period effect emerged for both mortality and DALYs. The relative risks showed gradual upward trends from 1992 to 2017 – increasing by 1.4-fold for mortality and 1.1-fold for DALYs – with the highest risks observed in 2017 (Table 2). Sex-specific analyses indicated that male risks exceeded female risks in 1992, 2002, 2007, and 2012 for both outcomes (Table S2). Accounting for age and period effects revealed a substantial cohort effect on both mortality and DALYs. Mortality risk was higher in earlier birth cohorts, with relative risks declining steadily from the 1897 to 1902 cohort to the 1992 to 1996 cohort (Table 2). For DALYs, risk peaked in the 1927 to 1932 cohort before declining thereafter. In terms of sex differences, male risks surpassed female risks from the 1897 to 1902 cohort through the 1947 to 1952 cohort for both mortality and DALYs (Table S2). 3.5. Cross-country inequality analysis Analysis of SHS-attributable lung cancer revealed significant SDI-related inequalities, which narrowed considerably over time (Fig. 5, Fig. S12). A disproportionate burden – encompassing both mortality and DALYs – was concentrated in countries with higher sociodemographic development. Between 1990 and 2021, the SII for ASDR decreased from 81 to 31 per 100,000 across the SDI spectrum, accompanied by a marginal increase in the CI from 0.21 (95% CI: 0.16–0.25) to 0.23 (95% CI: 0.17–0.30). Regarding ASMR, the corresponding SII disparity in ASMR declined from 3 per 100,000 to 1 per 100,000 over the same period, alongside a CI increase from 0.19 (95% CI: 0.14–0.23) to 0.24 (95% CI: 0.17–0.31). Together, these indices reflect a persistently unequal distribution of burden across countries at varying SDI levels. SDI-related health inequality regression (A) and concentration (B) curves for the ASDR of lung cancer attributable to secondhand smoke worldwide, 1990 and 2021. ASDR = age-standardized DALYs rate, SDI = sociodemographic index. 3.6. Predictive analysis on lung cancer attributable to SHS burden to 2035 Projections for lung cancer attributable to secondhand smoke (SHS) through 2035 are presented in Figure 6, with detailed estimates provided in Table S3. Worldwide, the absolute burden of deaths and DALYs is predicted to increase annually, set against a steady decline in their ASRs over the same timeframe. Global projections to 2035 for SHS-attributable lung cancer are illustrated in panels A–D, covering (A) mortality case numbers, (B) ASR of mortality, (C) DALY case numbers, and (D) ASR of DALYs. ASR = age-standardized rate, DALYs = disability-adjusted life-years, SHS = secondhand smoke. 4. Discussion The GBD 2021 database for the first time listed SHS as a Tier 3 risk factor and removed it from the PM2.5 category. This study is to report on the latest lung cancer mortality and DALYs attributable to SHS, comparing across regions, countries, sex, age groups, and time periods. Our analysis integrates trend, inequality, and predictive approaches to offer a comprehensive assessment. Over the period 1990 to 2021, a decline was observed in ASMR and ASDR for SHS-attributable lung cancer, whereas the corresponding absolute burden of deaths and DALYs exhibited an upward trend. With economic development and improved healthcare levels, the average age of the population is constantly increasing, and the aging population problem is becoming increasingly prominent. The above opposite trends may be due to changes in population structure, which is also corroborated by our observation that the burden of SHS-related lung cancer is heavier among the elderly. Numerous studies consistently show that SHS plays an important role as a risk factor for lung cancer progression.[17] A systematic review and meta-analysis showed that exposure to SHS increases the risk of lung cancer in never smokers by more than 20%.[22] As an environmental factor, SHS has a cumulative time effect in its carcinogenicity. Therefore, with the increase of cumulative effects, the number of deaths and DALYs caused by SHS-related lung cancer shows an increasing trend. The cumulative effect is also reflected in the age trend of the disease burden of SHS-related lung cancer. East Asia bore the greatest global burden of SHS‑related lung cancer mortality and DALYs in 2021, and its ASMR and ASDR levels remained the highest. We should not only raise concerns about regions with the highest current burden, but also about those where the increase or decrease has been greatest over the past few decades. Country‑specific and flexible health strategies are needed to address the marked disparities in the level and trajectory of SHS‑attributable lung cancer burden observed across regions. East Asia, as the most densely populated region globally, has the highest rates of SHS-attributable lung cancer deaths and DALYs, which may be related to its large population, just as China has the highest rates of SHS-attributable lung cancer deaths and DALYs at the national level. Smoking rates in low‑ and middle‑income countries increased from the early 1970s onward. By 2009, these nations collectively accounted for 82% of the world’s 1.1 billion smokers, with Asia alone representing more than half of this share.[23] As the world largest tobacco producer and consumer, China has approximately 300 million smokers, with males accounting for 96% of this figure.[24] According to the 2018 China Tobacco Control Report, approximately half of nonsmokers in homes or public places are still exposed to SHS.[25] By September 2010, over 170 countries had ratified national smoke-free legislation, a key policy under the World Health Organization Framework Convention on Tobacco Control.[26] The consistent decline in the ASMR and ASDR for SHS-attributable lung cancer across all SDI regions from 1990 to 2021 was likely related with the widespread implementation of tobacco control policies. The success of smoking ban implementation hinges largely on national attitudes and enforcement approaches, with developed countries tending to outperform developing ones according to available evidence.[27] Because developing countries may lack knowledge and awareness of the harmful effects of passive smoking, this limits support for smoking bans.[28] This also leads to differences in disease burden across different SDI regions. High SDI regions saw the largest declines in ASMR and ASDR, thanks to the early implementation of effective tobacco control measures and abundant healthcare resources in regions like Western Europe, North America, and Australia. Australia’s plain packaging law and strict tobacco advertising restrictions have served as effective tobacco control measures, leading to significant declines in smoking prevalence[29] In parallel, North America has strengthened its anti-smoking efforts by integrating public health campaigns with educational programs.[30] In 2021, the highest age-standardized mortality and DALY rates for SHS-attributable lung cancer were observed in high-to-medium SDI regions. This may stem from the confluence of 3 factors: historically high smoking prevalence in areas such as South America, Eastern Europe, and East Asia during the second half of the 20th century; the extended latency period characteristic of SHS-induced carcinogenesis; and pronounced socioeconomic disparities that heighten tobacco exposure among lower-income populations while limiting their access to early screening and quality medical care[31–33] Apart from China, the countries with the highest SHS-attributable lung cancer burden in 2021 were predominantly located in Eastern Europe and the Balkans. In these nations, smoking remains culturally entrenched, with high social acceptability sustained across generations,[34] owing to later adoption and less stringent enforcement of tobacco control policies relative to North America and Western Europe, exposure levels in these countries have remained persistently high.[33,35] Moreover, these countries have a significant gap in early cancer screening and diagnosis.[36] The National Lung Cancer Screening Trial was launched in the United States, the recent Cochrane review indicated that the mortality rate of lung cancer in high-risk populations who accept Low-Dose Computed Tomography screening has decreased by 20%.[37] National or regional lung cancer CT screening initiatives have been established across parts of North America, Europe, and Asia.[38,39] Nonetheless, the global uptake of such programs remains highly uneven, with resource constraints frequently hindering screening and novel therapeutic approaches in low- and middle-income countries. A higher SDI is typically associated with more robust implementation of preventive strategies. By comparison, in countries where age‑standardized mortality from lung cancer is low, the prioritization of lung cancer prevention and screening is often limited – likely because competing public health concerns are deemed more urgent.[40,41] The correlation between ASMR and ASDR of lung cancer caused by SHS in 2021 and SDI at regional and national levels also supports this view. Males have experienced a steeper decline in SHS-related lung cancer mortality and DALYs than females. Indeed, smoking prevalence is higher among men, whereas women and children are more likely to be exposed to SHS. Studies have found that in Southeast Asia, both overall and in each country, more women than men die from SHS related to tobacco.[42] The specific burden of SHS is higher among Southeast Asian women than the burden directly attributable to smoking.[43] It is noteworthy that a significant portion of lung cancer patients in developing countries are women, particularly nonsmokers. A Meta-Analysis of Epidemiologic Studies shows that SHS may increase the overall risk of lung cancer in never-smokers, especially among women.[44] Other studies show that approximately 16% of lung cancer cases in Chinese never-smokers may be attributed to passive smoking, with a slightly higher proportion (around 18%) in women. Most cases are caused by household exposure.[45] Unlike simply showing the trend of disease prevalence with age in a cross-sectional manner, to explore the independent effect of age on disease risk, an age-period-cohort model was applied in this study, controlling for other confounding factors, more clearly demonstrates that aging is one of the most prominent risk factors contributing to the increased burden of SHS-induced lung cancer. Middle to older age represents a convergence of factors – peak accumulation of SHS-induced lung damage and mutations, coupled with waning immune competence and repair capacity – which likely explains the disproportionate burden in this demographic. This aligns with previous evidence indicating that older and middle‑aged individuals carry the highest burden of smoking‑attributable tracheal, bronchus, and lung cancer deaths and DALYs.[46] The global aging trend underscores the need for more tailored preventive, management, and screening strategies targeting this age population. With the progression of the period, the relative risks for mortality and DALYs exhibited an increasing trend. One contributing factor is the upward trend in smoking rates among developing countries that started in the early 1970s, mentioned above,[23] The global population growth in the past 50 years has also been concentrated mainly in developing countries, compounded by the long latency period of SHS inducing lung cancer. Another contributing factor to the rising burden of SHS-related lung cancer is the gradual enhancement in health registration systems. In terms of birth cohort effects, as the cohort increases, the relative mortality rate and DALYs show a downward trend, meaning that individuals born earlier exhibit a higher risk of lung cancer attributable to SHS compared to those born later. Individuals born later have a clearer understanding of the harms of SHS compared to those born earlier, which is not only due to a growing understanding of the harms of SHS, but also the effectiveness of public health departments in preventing tobacco use, encouraging smoking cessation, and reducing SHS exposure among nonsmokers.[47,48] Quantifying SDI-related cross-national disparities in SHS-attributable lung cancer burden helps elucidate the pattern of disease distribution. The generally lower burden observed in high-SDI countries is often attributed to superior healthcare access and more comprehensive protective measures. However, an observed disproportionate burden of lung cancer attributable to SHS is concentrated in countries with higher socioeconomic development levels, which is consistent with our previous findings: at the national level, in 2021, ASMR and ASDR of lung cancer attributable to SHS were positively correlated with SDI. The inequality of SHS-induced lung cancer burden concentrating in high SDI countries may be mainly attributed to 2 aspects. Firstly, high SDI countries inevitably experience population aging. Secondly, in developed countries, the cumulative effect of a previously high smoking rate.[49] From 1990 to 2021, the SII and CI of lung cancer mortality and DALYs caused by SHS showed a decreasing trend, with both absolute and relative inequalities decreasing. This decrease may be attributed to global smoke-free legislation, stricter smoking bans in developed countries, and more advanced lung cancer screening technologies.[27,37] We still need targeted public health strategies to address these disparities and mitigate the impact of risk factors in different SDI environments. Although the ASMR and ASDR for SHS‑attributable lung cancer are expected to decline annually through 2035, the absolute burden of deaths and DALYs is projected to rise – a finding that underscores the enduring challenges in SHS hazard control and lung cancer screening management. The United Nations projects that global population will reach 9.7 billion by 2050, accompanied by accelerating aging.[50] From 2015 to 2050, the percentage of people aged 60 and over is forecast to nearly double, rising from 12% to 22%.[51] Population dynamics, including both growth and age‑related structural changes, as well as the cumulative effects of secondhand smoke, may be the reasons for the projected increase in lung cancer deaths and DALYs caused by secondhand smoke by 2035. The observed and projected increases in lung cancer deaths and DALYs attributable to SHS from 1990 to 2035 underscore the necessity for nations worldwide to strengthen tobacco control policies. These efforts are critical to mitigate the impact of SHS on nonsmokers and to prepare healthcare systems for demographic shifts, particularly through focused attention on aging populations and the strategic allocation of medical resources. However, several limitations need to be considered. Firstly, the GBD estimates for SHS‑attributable lung cancer may be subject to underestimation due to underdiagnosis, missed cases, and data loss in healthcare systems of less developed countries, compounded by inherent data gaps from heterogeneous country‑level inputs – particularly in regions such as sub‑Saharan Africa, although GBD collaborators sought to minimize these limitations through comprehensive data cleaning, correction, and large‑scale advanced statistical modeling. Secondly, our study did not consider the interaction between active and passive smoking, which could affect our findings. This may lead us to underestimate the burden of lung cancer caused by SHS. Furthermore, the lag in GBD data should be noted, as current estimates are calculated using past trends and covariates. Meanwhile, we must acknowledge that existing national cancer registries annually collect information on cancers with clear clinical diagnoses, which helps us understand the current global cancer progression and provides a basis for global public health policy development and resource allocation. To improve the accuracy of disease burden estimates, future global cancer research should focus on expanding international cooperation, implementing standardized cancer diagnostic criteria that include less developed countries, and strengthening cancer data collection in low-income nations. 5. Conclusions Although the total number of lung cancer deaths and DALYs attributable to SHS continued to rise from 1990 to 2021, both the ASMR and ASDR declined – a trend projected to persist through 2035. This pattern indicates a gradual reduction in the SHS-attributable lung cancer burden, though the burden remains disproportionately high among males and in regions with a high SDI. Collectively, these findings underscore the pressing need for governments to implement and enforce comprehensive smoke‑free legislation. In addition, they underscore the importance of establishing robust, continuous surveillance systems to monitor exposure to harmful substances. Acknowledgments The invaluable work of the Global Burden of Disease Study 2021 collaborators is gratefully acknowledged for providing the most comprehensive analysis of diseases worldwide. Author contributions Conceptualization: Xiao Wang, Haizhou Dang. Data curation: Xiao Wang, Yezu Liu, Liaoliao Bai, Hongqing Chen. Formal analysis: Xiao Wang, Yezu Liu, Liaoliao Bai, Hongqing Chen, Ailin Zhu. Investigation: Xiao Wang, Yezu Liu, Hongqing Chen, Ailin Zhu. Methodology: Xiao Wang, Liaoliao Bai, Ailin Zhu. Project administration: Yezu Liu, Haizhou Dang. Resources: Haizhou Dang. Software: Xiao Wang, Yezu Liu. Supervision: Yezu Liu, Haizhou Dang. Validation: Xiao Wang, Yezu Liu, Liaoliao Bai, Hongqing Chen. Visualization: Xiao Wang, Yezu Liu. Writing – original draft: Xiao Wang, Ailin Zhu. Writing – review & editing: Xiao Wang, Yezu Liu, Haizhou Dang.