Abstract
Introduction
This study investigates the cumulative association between adverse childhood experiences and lung cancer risk and analyzes whether a healthy lifestyle could modify this association.
Methods
Using the UK Biobank, we analyzed 156,798 participants who completed a baseline Mental Health Questionnaire. Lung cancer risk from age 30 was recorded, classifying childhood adversity severity using cumulative categories (none, mild [1-2 types], and severe [≥3 types]). We used multi-variable Cox regression and competing risk models to estimate hazard ratios (HR) between childhood adversity and lung cancer risk. Mediation analyses assessed the role of smoking.
Results
During a median follow-up of 41 years from age 30, 677 participants were diagnosed with lung cancer, and 401 died. Lung cancer risk increased progressively with increasing childhood adversity severity. Compared with individuals without childhood adversity, those who experienced any adversity had a 37% higher risk of lung cancer (HR 1.37, 95% CI 1.15-1.63), while those with severe adversity had an 82% higher risk (HR 1.82, 95% CI 1.27-2.62), demonstrating a clear summative effect. Participants who experienced any form of adversity had a 18% increased risk of dying from lung cancer (HR 1.18 [1.05-1.32]). Smoking severity was higher among participants with childhood adversity and accounted for approximately 40% of the observed association. Timely smoking cessation was associated with a substantially lower risk of lung cancer (approximately 75%). However, in the severe adversity group, smoking cessation did not provide a protective effect (HR 0.47 [0.20-1.09]). Smoking cessation significantly reduced lung cancer mortality regardless of adversity level.
Conclusion
Childhood adversity was associated with a higher risk of lung cancer in a cumulative manner, even after accounting for non-cancer death. Smoking accounted for a substantial proportion of the association between childhood adversity and lung cancer risk, and smoking cessation was associated with lower lung cancer risk across adversity levels. The association between childhood adversity and lung cancer remained robust, warranting further research to explore additional mediating pathways.
Keywords
Introduction
Lung cancer is one of the most common cancers worldwide and is also the leading cause of cancer-related deaths, causing millions of deaths each year. 1 In addition to what we already know, poor lifestyle habits, such as smoking, are directly associated with the development of lung cancer as independent risk factors.2-4 Current research is gradually showing that support for the link between psychosocial factors in cancer development continues to grow. Biobehavioural factors not only influence cellular immunity, but also directly or indirectly regulate fundamental processes of cancer growth, including inflammation, angiogenesis, invasion and metastasis.5,6
Childhood adversity is defined as children experiencing physical or emotional abuse or neglect, which can be a major source of stress and have short- or long-term effects on their health. 7 Numerous studies have shown that adverse childhood experiences (ACEs) have long-term effects on health throughout the life course and that exposure to ACEs is associated with an increased risk of mental illness, respiratory disease, cardiovascular disease, and cancer.2,8 ACEs also led to increased health-damaging behaviors such as smoking, harmful alcohol use, and drug use. 8 Biopsychosocial mechanisms may be involved in how ACEs affect disease development. Some studies suggest that ACEs trigger hyperactivation of the hypothalamic-pituitary-adrenal axis and the sympathetic and parasympathetic nervous systems. Increased levels of angiotensin-converting enzyme directly affect hormonal and immune development.9,10
Although previous studies have reported an association between ACEs and increased risk of cancer, few have focused on any specific type such as lung cancer,2,11 nor did their potential mediators. Based on that, we conducted a population-based prospective cohort study using the UK Biobank (UKB) database to investigate the association between childhood adversity and lung cancer incidence and mortality. We also analyzed whether these associations could be modified by adopting a healthy lifestyle. Through this study, we hope to provide a new rationale for public health and clinical interventions to help develop more effective prevention strategies to reduce lung cancer incidence and mortality.
Method
Study Design and Population
This study utilized the UK Biobank database, a prospective cohort study that recruited over 500,000 participants from 22 assessment centers across England, Scotland, and Wales between 2006 and 2010.
12
Of these participants, 157,239 completed an online questionnaire on adverse childhood experiences. After excluding those with incomplete questionnaires and incomplete inclusion characteristics, 156,798 participants were included in the analysis (Figure 1). All participants provided informed consent. Ethical approval for this study was obtained from the North West-Haydock Research Ethics Committee (Reference: 21/NW/0157) through UK Biobank. Study selection
Assessment of Childhood Adversity
Each participant’s childhood adversity was assessed using the web-based Mental Health Questionnaire, developed to assess adverse childhood experiences. The questionnaire asked whether participants had experienced physical abuse, emotional abuse, sexual abuse, emotional neglect, or physical neglect during their childhood. Each type of adversity was rated on a 5-point scale from “never” to “often” (scored from 0 to 4). Although this scale allows a more granular characterization of adversity severity, severity-based analyses at the individual item level were considered exploratory, as higher severity categories were uncommon and associated with limited statistical power. Therefore, consistent with prior population-based studies, we primarily operationalized childhood adversity based on the presence of each adversity type rather than frequency alone. Exploratory severity-based analyses of individual childhood adversity domains and lung cancer risk are presented in Supplemental Table 3. The presence of each type of adversity was defined as follows: emotional neglect and physical neglect were scored ≤1, while physical, emotional, and sexual abuse were scored ≥2. Participants who reported at least one of the five types of adversity were categorized as having experienced childhood adversity, while those who reported none were categorized as not having experienced childhood adversity. Participants were further categorized into three groups based on their childhood adversity: no childhood adversity, mild childhood adversity (one to two adversities), and severe childhood adversity (three or more adversities).13,14
Lung Cancer Incidence and Mortality
The primary endpoints of this study were lung cancer incidence and mortality, as assessed by ICD-9 (162) and ICD-10 (C33, C34) hospitalization and mortality data. The follow-up period for each participant was defined as the time between the age of 30 years and the occurrence of lung cancer (or death from lung cancer) or the end of follow-up (September 1, 2023). Follow-up was initiated at age 30 because lung cancer occurrence before this age is extremely rare, and the present study focused on adult-onset lung cancer risk. In addition to lung cancer incidence, lung cancer-specific mortality was analyzed as a survival-related outcome.
Covariates
Baseline characteristics collected for each participant were included in the analysis as covariates. These included age, sex, race, Townsend Deprivation Index (TDI), smoking history, and family history of cancer. The proportion of missing data for covariates was typically less than 1%. We employed Multiple Imputation by Chained Equations (MICE) to handle missing data. This method generates multiple imputations by iteratively considering the chained relationships among variables during the imputation process. MICE ensures both flexibility and stability in the imputation procedure. 15 Given the low proportion of missing data across all covariates, no separate formal evaluation of imputation performance using a fully observed subsample was conducted. Instead, the robustness of the analyses was assessed through a series of sensitivity analyses, including additional covariate adjustments and competing risk models, which yielded results consistent with the primary findings. Smoking status was assessed at baseline and categorized as current or former smoking. Former smokers were defined as individuals who had quit smoking prior to cohort entry, which we refer to as timely smoking cessation. No specific threshold for time since quitting was applied due to substantial missingness and potential recall bias.
Statistical Analysis
We used Cox proportional hazards models to estimate the association between childhood adversity and lung cancer incidence or mortality, adjusting for age, sex, race, TDI, smoking history, and family history of cancer. In addition to smoking status, smoking severity was explored using cumulative smoking exposure measured as pack-years among participants with available data (Supplemental Table 5). Pack-year smoking was evaluated as a continuous variable in Cox proportional hazards models. Interaction terms between pack-year smoking and childhood adversity were examined to assess potential effect modification. Results are reported as hazard ratios (HRs) with 95% confidence intervals (95%CI). The p-values were adjusted using the false discovery rate (FDR), and the proportional hazards assumption was tested using the Schoenfeld residuals test. Childhood adversity was categorized into three groups: no childhood adversity, mild and severe childhood adversity, and the no adversity group regarded as the reference. We also analyzed the effect of each of the five types of adversity on the outcomes.
Regarding the model diagnostics, age was found to violate the proportional hazards assumption and was therefore modeled as a stratification variable in all Cox regression analyses. After stratification by age, the global Schoenfeld test did not indicate a significant violation of the proportional hazards assumption. Although childhood adversity showed a mild deviation from the proportional hazards assumption at the individual covariate level, sensitivity analyses using time-varying Cox models indicated that this deviation did not materially affect the overall conclusions. Accordingly, hazard ratios from the primary model were interpreted as average effects over the follow-up period.
We performed sensitivity analyses to assess the robustness of the association between childhood adversity and lung cancer. First, we adjusted our model for alcohol status, BMI, CRP (an inflammatory marker), and comorbidities including hypertension, hyperlipidemia, diabetes, cardiovascular disease, anxiety, and depression. Diagnoses of hypertension, hyperlipidemia, and diabetes were based on baseline biochemical and clinical indicators, whereas diagnoses of cardiovascular disease, anxiety, and depression were derived from ICD-10 codes, considering only those comorbidities occurring before lung cancer diagnosis. Finally, we used Fine-Gray proportional hazards models to account for competing risks of non-lung cancer deaths.
Mediation analyses were performed to examine whether the association between childhood adversity and lung cancer risk was mediated by smoking. Logistic and linear models were constructed to regress exposure (childhood adversity) on the mediating variable, and survival models were used to regress the exposure and mediator on lung cancer, adjusting for potential confounders. These regression models were integrated, and the “mediator” R package was used to estimate direct, indirect, and mediated effects proportions.14,16-18 Hazard ratios and 95% CIs were calculated to evaluate the effect of smoking cessation among those who experienced childhood adversity.
No formal a priori sample size calculation was performed, as this study represents a secondary analysis of an existing population-based cohort (UK Biobank) with a fixed sample size.
Results
Baseline Characteristics of the Study
The Basic Characteristics of the Study
Association Between Childhood Adversity and Lung Cancer Incidence
During a median follow-up of 41 years from age 30, a total of 677 subjects were diagnosed with lung cancer, of which 401 died. Among the lung cancer patients, 184 had experienced any form of adversity, with 33 having experienced severe childhood adversity. Participants who experienced adversity had a 37% increased risk of developing lung cancer compared to those who did not experience childhood adversity (HR: 1.37 [1.15, 1.63]). Severe adversity was associated with an 82% higher risk of lung cancer (HR 1.82 [1.27, 2.62]), which was higher than the risk associated with mild adversity (HR 1.27 [1.06, 1.53]). Different types of adversity were also linked to an increased risk of lung cancer, with childhood sexual abuse having the highest association, increasing the risk by 74%. Following sexual abuse, physical neglect was associated with a 56% increased risk (HR 1.56 [1.12, 2.18]). Emotional abuse, physical abuse, and emotional neglect were all associated with more than a 30% increased risk of developing lung cancer (Figure 2; Supplemental Tables 1 and 2). When we additionally adjusted the model for alcohol consumption, BMI, CRP, or comorbidities, the effects of sexual abuse and emotional neglect on lung cancer incidence were not statistically significant, while the other results remained stable (Supplemental Table 6). The results from the competing risk analysis, which considered other causes of death, were robust (Supplemental Table 8). The correlation between childhood adversity and lung cancer incidence
Association Between Childhood Adversity and Lung Cancer Mortality
Lung cancer deaths occurred in 112 people who experienced any childhood adversity and in 16 people who experienced severe childhood adversity. The incidence and mortality of lung cancer were twice as high in the severe adversity group compared to the non-adversity group. For lung cancer mortality, subjects who experienced adversity had an 18% increased risk of dying from lung cancer (HR 1.18 [1.05, 1.32]).When stratified by severity, mild adversity was associated with a 40% higher risk, whereas severe adversity showed a 51% higher hazard, although this estimate did not reach statistical significance (HR 1.51 [0.89, 2.55]). In the analysis of different types of adversity, emotional abuse (HR 1.43 [1.06, 1.93]) and physical abuse (HR 1.36 [1.00, 1.87]) were significantly related to lung cancer mortality (Figure 3; Supplemental Tables 1 and 2). When we additionally adjusted the model for alcohol consumption, BMI, CRP, or comorbidities, the effect of any childhood adversity and mild adversity on lung cancer mortality remained significant (Supplemental Table 7). Emotional abuse was statistically significant for lung cancer mortality, and its clinical relevance also warrants attention (HR 1.43 [1.06, 1.93]). The results from the competing risk analysis, which included other causes of death, were robust (Supplemental Table 8). The correlation between childhood adversity and lung cancer mortality
Mediation Analysis of Smoking and Smoking Cessation Effectiveness
Smoking showed a substantial contribution to the association between childhood adversity and lung cancer risk, motivating further examination of whether smoking cessation was associated with differences in this relationship. Mediation analyses suggested that approximately 39.0% of the observed association between childhood adversity and lung cancer risk could be statistically attributed to smoking, with a natural direct effect estimate of HR 1.32 (95% CI 1.11, 1.57). There were no significant mediating effects for other variables (Supplemental Figure 9-19). Similarly, for lung cancer deaths, smoking was a significant mediating variable, with 40.9% of the effect mediated by smoking and a natural direct effect of HR 1.38 (1.10, 1.72) (Supplemental Figure 20-30). To further understand the association between childhood adversity and lung cancer risk, it was assessed by analyzing groups of never, former and current smokers, using current smokers as the reference for each adversity group.
When examining smoking cessation as a smoking-related behavioral factor, we observed that smoking cessation was associated with more favorable outcomes across different adversity groups (Figure 4). Here, smoking cessation was defined as having quit smoking prior to baseline assessment, rather than continued smoking, without imposing a specific duration since cessation. For lung cancer development, never-smokers who experienced any childhood adversity had a 93% reduced risk compared with those who still smoked (HR 0.07 [0.05, 0.12]), while timely smoking cessation also conferred a significant benefit, although slightly less than those who had never smoked (HR 0.25 [0.18, 0.34]). The health behaviors of never smoking and immediate quitting were equally beneficial for lung cancer development in the mild adversity group, with HRs of 0.06 (0.04, 0.10) and 0.22 (0.15, 0.31), respectively. However, with increasing severity of childhood adversity, the magnitude of the inverse association between non-smoking and lung cancer risk appeared to attenuate. Among participants with severe adversity, those who had never smoked showed a substantially lower risk of lung cancer (HR 0.19 [0.06, 0.55]), whereas among individuals with severe adversity who were former smokers, smoking cessation was not significantly associated with a lower risk of lung cancer (HR 0.47 [0.20, 1.09]). Effectiveness of smoking cessation in reducing lung cancer incidence risk among individuals with childhood adversity
In contrast, smoking cessation was effective in reducing lung cancer deaths in all populations. Never smoking reduced the risk of lung cancer death by 96% in those who experienced childhood adversity (HR 0.04 [0.02, 0.08]), and smoking cessation also reduced the associated risk by 81%. The effect was similar for those with mild adversity. In the severe adversity population, non-smoking was a stronger protective factor against lung cancer death, reducing the risk by 97% compared with the smoking population. Smoking cessation remained effective in reducing lung cancer deaths (HR 0.17 [0.06, 0.52]). It is important to note that smoking itself is a major risk factor for lung cancer death. However, for those who have experienced childhood adversity, the benefits of quitting smoking are even greater (Supplemental Table 4).
Discussion
This study examined the association between childhood adversity and lung cancer incidence and mortality in adulthood through a population-based cohort study of over 150,000 individuals. The results not only showed that childhood adversity was significantly associated with lung cancer but also revealed a positive association between the severity of adversity and the risk of lung cancer incidence and mortality. Smoking was identified as a key contributor to the observed association between childhood adversity and lung cancer incidence and mortality. Moreover, smoking cessation was associated with a reduced risk of lung cancer and was effective in populations with different levels of childhood adversity.
Previous studies are consistent with our findings and suggest that childhood adversity is strongly associated with the development of respiratory disease and cancer.2,11,19 The current study provides preliminary evidence of the association of ACE with respiratory disease, highlighting the increased risk of long-term non-communicable diseases. 20 However, previous studies have focused primarily on the association of ACEs with asthma and COPD.13,19 Meanwhile, a meta-analysis of four studies found that ACE was associated with cancer, conferring a more than twofold increased risk of cancer.2,21 Despite this, previous cohort studies of ACE have been small, and cancer has often been reported as one of the general outcomes, without clarifying the relationship between ACE and specific cancers and the possible mediators.11,20 Additionally, no study has focused specifically on the association between childhood adversity and lung cancer.
The results of our study showed that people who experienced any form of childhood adversity had a 37% increased risk of developing lung cancer, while those who experienced severe adversity had a 79% increased risk. There was also a 41% increased risk of dying from lung cancer. Exposure to adverse childhood experiences is a risk factor for future development, particularly during developmentally sensitive periods.8,22 These findings further support the long-term effects of childhood adversity on adult health. Further analysis showed that childhood sexual abuse, physical neglect, and emotional abuse were the types of adversity that had the greatest impact on the risk of developing lung cancer, with sexual abuse showing a particularly significant 74% increase in risk. These significant increases in risk suggest that specific childhood adversity experiences may have a profound impact on an individual’s long-term health through complex biopsychosocial mechanisms, particularly in the development of lung cancer.
Our study found that almost 10 percent of the population had experienced emotional or physical aggression and abuse during childhood, and more seriously, 4.11 percent had experienced sexual abuse. There was a higher percentage of women in the severe adversity group, suggesting that women are more likely to be victims of adverse childhood experiences, or that such experiences are more deeply remembered and have a broader impact on them. 23 Nevertheless, the proportion of smokers increased progressively with the severity of adversity. Consistent with our findings, previous research has shown that ACE is associated with increased smoking behavior across all regions.2,21 These findings may reflect different coping mechanisms adopted by individuals in response to adversity, such as increased risk behaviors like smoking. 11
Using mediation analysis, we found that smoking played a significant mediating role between childhood adversity and lung cancer incidence and death, with 39.5% and 40.9% of the associations mediated by smoking, respectively. These findings extend our primary analyses by demonstrating that not only smoking status but also smoking intensity contributes to lung cancer risk, supporting a dose-dependent effect of tobacco exposure. Together, this results highlights the important role of smoking as a coping mechanism for adversity and suggests the potential effectiveness of smoking cessation interventions in lung cancer prevention. Despite the significant effect of smoking on lung cancer mortality, the protective effect of smoking cessation in reducing lung cancer mortality was significant in all populations. Non-smokers had a significantly lower risk of lung cancer than smokers after experiencing any form of childhood adversity, and the protective effect of non-smoking and timely cessation on the development of lung cancer was particularly significant in the mild adversity group. It should be noted that smoking cessation did not statistically significantly reduce lung cancer development in the severe adversity group, probably because of the strong association between severe childhood adversity and lung cancer development, which weakened the protective effect of smoking cessation. However, we believe that the clinical importance of smoking cessation remains, as it plays an important role in reducing lung cancer incidence and deaths in those who have experienced childhood adversity.
Clinical Implications
Although the mechanisms by which childhood adversity contributes to the development of lung cancer remain incompletely understood, childhood adversity may influence lung cancer development through multiple biopsychosocial mechanisms.9,24,25 In this study, smoking emerged as an important explanatory factor in the association between childhood adversity and lung cancer. Notably, smoking cessation and never smoking were associated with lower lung cancer risk even among individuals who had experienced childhood adversity, highlighting smoking behavior as a potentially modifiable factor in this high-risk population. Given the observational nature of this study, future research using causal inference methods, such as Mendelian randomization, is warranted to further assess the potential causal pathways linking childhood adversity, smoking, and lung cancer risk.26-28 Previous research has linked adverse childhood experiences to a broad range of noncommunicable diseases, yet prevention efforts have largely focused on downstream behavioral risk factors. 11 Further studies are needed to evaluate whether interventions targeting childhood adversity could complement existing prevention approaches. 29
Study Limitations
Our study inevitably has some limitations. First, recall bias may affect the accuracy of reported childhood adversity, leading to imprecise results. Secondly, as a cohort study, our study has a relatively limited ability to interpret causality, which reduces the level of evidence. Despite our efforts to adjust for potential confounders, the influence of some residual confounders could not be completely excluded. In addition, UK Biobank participants were predominantly recruited at middle to older ages (40–69 years). As a result, individuals who developed and died from lung cancer before cohort recruitment were not captured. This survivor selection may limit the generalizability of our findings to very early-onset lung cancer and could potentially lead to an underestimation of the observed associations if childhood adversity increases the risk of early fatal disease. This left-truncated cohort design, inherent to UK Biobank recruitment, may preferentially include individuals who survived to middle age, thereby potentially attenuating observed associations. Furthermore, the classification of adversity severity in our study was based solely on the number of self-reported adversity types. However, even a single type of adversity can cause severe psychological trauma in children. Therefore, further assessment using more elaborate scales is warranted in future studies. In addition, the vast majority of the population included in this study was white and European, so the findings need to be interpreted and generalized with caution.
Conclusion
Our study, utilizing large-scale prospective cohort data, reveals a significant association between childhood adversity and lung cancer risk and emphasizes the importance of different types and severity levels of childhood adversity in lung cancer incidence and mortality. Smoking cessation was associated with a substantial attenuation of the elevated lung cancer risk and mortality observed among individuals exposed to childhood adversity. The study provides preliminary evidence suggesting potential pathways linking childhood adversity to lung cancer development and mortality through unhealthy behaviors such as smoking. Future research should employ advanced techniques, such as Mendelian Randomization, to validate these findings and further elucidate the causal pathways. Additionally, adverse childhood experiences should be considered in future lung cancer interventions to formulate more effective prevention strategies and public health policies, thereby mitigating their long-term impact on adult health.
Supplemental Material
Supplemental Material - Impact of Adverse Childhood Experience on Lung Cancer Risk: A Population-Based Prospective Cohort Study
Supplemental Material for Impact of Adverse Childhood Experience on Lung Cancer Risk: A Population-Based Prospective Cohort Study by Yu Zhang, Shen Li, Chi Shu, Siheng Tao, Guohua Shen, Xuan Xie, Chuan Xu, Jiyan Liu, Yazhou He, Hu Liao and Xuelei Ma in Cancer Control.
Footnotes
Ethical Considerations
Ethical approval for this study was obtained from the North West-Haydock Research Ethics Committee (Reference: 21/NW/0157) through UK Biobank.
Consent to Participate
All participants provided written informed consent at enrollment in the UK Biobank.
Author Contributions
Study concept and design: MXL, ZY, LS, SC, LH, HYZ, LJY, XC; Data analysis: ZY, LS, SC, HYZ, LJY, TSH; Visualization: ZY, LS, TSH, XX; Drafting of the Manuscript: MXL, ZY, LS, SC, HYZ, SGH. All authors read and approved the final manuscript.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study is funded by the Natural Science Foundation of Sichuan Province (grant no. 2025ZNSFSC0553).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data used in this study are available from the UK Biobank upon reasonable request and subject to approval by the UK Biobank. The analytical code and summary-level results supporting the findings of this study are available from the corresponding author upon reasonable request.
Supplemental Material
Supplemental material for this article is available online.