The impact of adolescent smoking initiation on the risk of end-stage kidney disease: a nationwide cohort study
Article information
Abstract
Background
Chronic kidney disease (CKD) is a growing global health challenge, with smoking identified as a significant risk factor. This study investigates the long-term impact of adolescent smoking initiation on end-stage kidney disease (ESKD) development.
Methods
A retrospective cohort study was conducted using data from the Korean National Health Insurance Service claims database. The cohort included 201,678 CKD patients aged ≥40 years with a documented smoking history. Patients were stratified by smoking initiation age (<20 years vs. ≥20 years) and cumulative smoking exposure (pack-year, PY). The primary outcome was ESKD incidence, defined as kidney replacement therapy initiation. Cox proportional hazards models assessed the relationship between smoking initiation age, smoking burden, and ESKD risk.
Results
During a median 6.8-year follow-up period, 6,334 patients progressed to ESKD (incidence rate, 3.63 per 1,000 PYs). Those with higher PYs were older and had more comorbidities, such as hypertension and diabetes mellitus. Patients who began smoking before age 20 years and accumulated ≥20 PYs had a significantly higher risk of ESKD (hazard ratio, 1.26; 95% confidence interval, 1.16–1.38) compared to those with the same exposure but later smoking initiation. Increased cumulative smoking exposure further elevated the risk. When PYs were divided according to initiation age, a higher ratio was associated with an increased risk of ESKD.
Conclusion
Early smoking initiation during adolescence was associated with a significantly higher risk of progression to ESKD in patients with CKD, especially in those with higher cumulative smoking exposure. Public health interventions focusing on preventing adolescent smoking can mitigate the long-term burden of CKD progression.
Introduction
The global increase in teenage smoking poses a significant public health concern. A survey of 12,328 adolescents in 11 European countries showed that 58% began smoking before the age of 14, highlighting the connection between early smoking initiation and a range of issues, including mental health problems, behavioral difficulties, and strained family relationships [1]. Additionally, evidence suggests that teenage smoking is not only associated with immediate health risks but also contributes to the development of long-term conditions, such as cardiovascular diseases, respiratory illnesses, and cancer [2–4]. Furthermore, adolescents exposed to tobacco smoke face higher risks of respiratory issues, poorer overall health, increased school absenteeism, and more frequent emergency department visits [5].
Simultaneously, chronic kidney disease (CKD) has emerged as a growing global health challenge, [6] with rising incidence and prevalence rates leading to increased morbidity and mortality [7–10]. This places CKD among the top public health priorities. In recent years, smoking has been increasingly recognized as a significant risk factor for the development and progression of CKD, with several studies demonstrating a dose-response relationship between smoking intensity and declining kidney function [11–14]. Given the substantial burden CKD places on healthcare systems, identifying and controlling modifiable risk factors, such as smoking, is crucial to mitigating its impact.
While the relationship between smoking and adverse health outcomes, including CKD, is well-established, the specific long-term effects of smoking initiation during adolescence on kidney health remain underexplored. This study aims to fill this gap by utilizing extensive data from the Korean National Health Insurance Service (NHIS) claims database to investigate how the age of smoking initiation, as well as the dose and duration of smoking, influences kidney health in patients with CKD. Specifically, we focus on the risk of developing end-stage kidney disease (ESKD) later in life. This research is essential for understanding the impact of early-life behaviors on chronic disease outcomes and for informing public health interventions aimed at reducing these risks.
Methods
Ethical considerations
This study was approved by the Institutional Review Board of the Seoul National University Hospital (No. E-2112-048-1281). Access to the Korean NHIS database was granted by the appropriate governmental authorities. The study adhered to the ethical guidelines outlined in the Declaration of Helsinki. Due to the retrospective nature of the study and the use of completely anonymized and untraceable data, the requirement for informed consent was waived. This study adhered to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines [15].
Data source
This study utilized the NHIS claims database, which includes detailed demographic data, healthcare usage, medical procedures, prescriptions, health examination results, and mortality statistics for all Korean residents. The NHIS covers 97% of the population and provides annual or biennial health screenings to over 10 million people, representing more than 20% of the total population. These comprehensive records were accessible through authorized research use [16,17].
Within this dataset, demographic characteristics and lifestyle habits were systematically collected using structured self-report questionnaires. A key variable included in the study was smoking history, which was initially categorized into three groups: never-smokers, ex-smokers, and current smokers, based on recorded smoking data.
To gain further insight, we stratified both ex-smokers and current smokers. This included a detailed assessment of daily cigarette consumption, categorized into less than one pack, one to two packs, or more than two packs per day, and smoking duration, classified as less than 10 years, 10–20 years, or >20 years. Cumulative smoking exposure was then measured in pack-years (PY) and classified into four groups: less than 10 PY, 10 to 20 PY, 20 to 40 PY, and 40 or more PY.
Study population
We initially screened individuals aged 40 years or older with CKD, defined as an estimated glomerular filtration rate (eGFR) of less than 60 mL/min/1.73 m2 or albuminuria detected by dipstick at 1+ or higher, based on national health examinations conducted by the NHIS between January and December 2009. Serum creatinine levels used to calculate eGFR were obtained from the NHIS health examination database. Creatinine was measured using an enzymatic method or the Jaffe method, depending on the institution, and all measurements were calibrated to isotope-dilution mass spectrometry traceable standards to ensure consistency across facilities. The eGFR was calculated using the MDRD (Modification of Diet in Renal Disease) equation. Patients who were never smokers at the time of screening or had missing data for any study variables were excluded. To minimize confounding, we designated the period from 2002 to 2009 as a “washout” period and excluded patients diagnosed with ESKD during this time. For analysis, patients were categorized into four groups based on two criteria: age at smoking initiation (under 20 years vs. 20 years or older) and cumulative PY exposure (up to 20 PYs vs. more than 20 PYs).
Outcome
The primary outcome of our study was the development of ESKD, defined as the initiation of kidney replacement therapy, which includes both dialysis and transplant procedures as captured by NHIS records. The outcome data collection began in 2009, with the follow-up period ending on December 31, 2018.
Covariates
The study incorporated a broad range of data, including demographic characteristics, lifestyle factors, initial kidney function, and various metabolic indicators. Data collected included the patient’s age, sex, body mass index (BMI), and alcohol consumption (any amount exceeding 0 g/day). Regular exercise habits were categorized as moderate-intensity activity for more than 5 days a week or vigorous-intensity activity for more than 3 days a week. Economic status was assessed by focusing on individuals in the lower quartile of national income. Medical history encompassed diabetes mellitus, identified using International Classification of Diseases, 10th Revision (ICD-10) codes E11–14, including diabetes medication, and hypertension, diagnosed with ICD-10 codes I10–13 or I15, along with relevant medication history. Initial kidney function was evaluated through eGFR and the presence of dipstick albuminuria (≥1+). Metabolic syndrome (MetS) was diagnosed if three or more of the following criteria were met during a health check-up: waist circumference exceeding 90 cm for male and 80 cm for female; elevated triglycerides (≥150 mg/dL) or use of related medication; decreased high-density lipoprotein cholesterol (<40 mg/dL for male and <50 mg/dL for female) or use of relevant medication; high blood pressure (systolic, ≥130 mmHg and/or diastolic, ≥80 mmHg) or use of antihypertensive medication; and elevated fasting glucose levels (≥100 mg/dL) or use of antidiabetic drugs. All baseline covariates were extracted from the NHIS claims data at the time point when CKD was defined.
Statistical analyses
Baseline characteristics were reported as means ± standard deviations for continuous variables and as numbers (percentages) for categorical variables. Continuous variables were compared using analysis of variance, while categorical variables were analyzed with the chi-square test. The incidence of ESKD was computed by dividing the number of events by 1,000 person-years. Multivariate Cox proportional hazards regression analyses were conducted to assess the association between smoking parameters and the development of ESKD. Results are reported as hazard ratios (HRs) with 95% confidence intervals (CIs). Adjustments were made for covariates including age, sex, income, alcohol consumption, exercise, BMI, eGFR, MetS, and proteinuria. These adjustments were consistently applied in subgroup analyses based on the presence or absence of MetS, diabetes mellitus, and hypertension.
To explore the association between age at smoking initiation and kidney outcomes, total PYs were stratified at cutoff points of 20, 30, and 40 years. Participants were further categorized based on smoking initiation: before 20 years of age vs. 20 years or older. To provide a more conservative assessment of statistical significance in subgroup comparisons, post-hoc analyses using Bonferroni-adjusted p-values were performed based on HRs derived from the Cox regression model, with a single reference group (“<20 PY and ≥20 years”) maintained.
To account for the potential inflation of PY values due to earlier smoking initiation, we adjusted for age at smoking initiation by creating a new variable that divides the total PY by the age at which smoking began. This approach accounts for both cumulative exposure and its accumulation over time, aligning with previous findings that prolonged, lower-intensity smoking may pose greater health risks than short-term, high-intensity smoking [18]. This variable was then categorized into quartiles. Additionally, participants were classified by their age at smoking initiation independently of their PY values, with PY used as an adjusting variable to explore trends in ESKD risk related to the timing of smoking initiation.
Results
Baseline characteristics
After applying the exclusion criteria, a total of 201,678 CKD patients with a history of smoking were included in the study from the original pool of 789,207 screened individuals (Fig. 1). Participants with higher PY were older and had a greater prevalence of comorbidities, such as hypertension, dyslipidemia, diabetes mellitus, and MetS, compared to those with lower PY. Notably, within the same PY category, participants who started smoking at a younger age exhibited higher rates of heavy alcohol consumption, elevated triglyceride levels, and an increased incidence of proteinuria. However, comorbidities such as hypertension, diabetes mellitus, and MetS were less prevalent in this group. Detailed demographic and clinical characteristics, including age, sex, and comorbidities, are shown in Table 1.
Study flow diagram.
CKD, chronic kidney disease; ESKD, end-stage kidney disease.
Baseline characteristics
Kidney outcomes according to smoking intensity stratified by initiation age
We investigated the association between smoking initiation age and ESKD development across various smoking exposure levels. Among individuals with fewer than 20 PYs of smoking, there was no significant increase in ESKD risk, regardless of whether smoking began before the age of 20 years. However, for those with more than 20 PYs, participants who started smoking earlier had a significantly higher risk of ESKD (adjusted HR [aHR], 1.26; 95% CI, 1.16–1.38). The increasing risk was more pronounced with higher levels of smoking exposure, as indicated by non-overlapping CIs. When the smoking burden threshold increased, the aHR for individuals who started smoking before age 20 years rose from 1.26 to 1.37. This trend persisted when the data were stratified by higher thresholds of smoking duration, such as 30 and 40 PYs (Table 2).
Risk of ESKD according to smoking intensity stratified by initiation age
Kidney outcomes according to age-standardized smoking intensity
We found that a higher age-standardized smoking intensity (PYs divided by smoking initiation age) was associated with a greater risk of ESKD. Fig. 2 illustrates the Kaplan-Meier curves for ESKD risk across quartiles of smoking intensity, showing an upward trend in the cumulative incidence from the lowest to the highest quartile. Multivariable Cox proportional hazards analysis confirmed this trend, with individuals in the highest quartile exhibiting a significantly elevated ESKD risk compared to those in the lowest quartile (aHR, 1.14; 95% CI, 1.06–1.22). The detailed risk assessments and CIs are presented in Table 3.
Kaplan-Meier survival curve for the development of ESKD according to quartile ranges of age-standardized smoking intensity, obtained by dividing pack-year by the age at smoking initiation.
The Y-axis represents incidence probability. Q4 represents the highest smoking intensity.
ESKD, end-stage kidney disease.
Risk of ESKD according to age-standardized smoking intensity
Association between smoking initiation age and the development of end-stage kidney disease
When stratifying the population by smoking initiation age, we found a progressively increasing risk of ESKD among those who began smoking at younger ages, even after adjusting for total PYs. Individuals who started smoking at age 25 years, 20 years, 17 years, and younger than 15 years all exhibited higher risks compared to those who initiated smoking at age 30 years or older. Fig. 3 highlights these trends, which remained consistent after adjusting for PYs to account for possible confounding by smoking intensity. The corresponding statistical data are shown in Table 4.
Risk of end-stage kidney disease according to smoking initiation age.
Model 1: adjusted for age, sex, income, drinking, exercise, body mass index, glomerular filtration rate, metabolic syndrome, and proteinuria. Model 2: adjusted for variables in model 1 plus pack-years.
CI, confidence interval.
Risk of ESKD according to smoking initiation age
Kidney outcomes according to smoking intensity stratified by initiation age: insights from subgroup analysis
In subgroup analyses, we found that smoking intensity was associated with higher ESKD risk among participants who started smoking before age 20 years and did not have MetS (aHR, 1.783; 95% CI, 1.120–2.840) or comorbidities such as diabetes mellitus (aHR, 1.497; 95% CI, 1.044–2.146) or hypertension (aHR, 4.608; 95% CI, 2.026–10.483) (Supplementary Fig. 1, available online). These findings suggest that early smoking initiation exerts a significant, independent effect on kidney disease progression, even in the absence of common metabolic risk factors.
Discussion
Our study presents significant findings on the relationship between early smoking initiation during adolescence and the progression to ESKD in patients with CKD. We found a marked increase in the risk of ESKD among individuals who began smoking before the age of 20 years, particularly in those with higher cumulative smoking exposure, measured in PY. Specifically, our results show that participants who initiated smoking during adolescence had a significantly higher aHR for ESKD compared to those who started smoking later in life, even with comparable smoking intensities, such as accumulating 20 or more PYs.
Our findings extend the existing body of research on smoking and CKD progression. Prior studies have consistently demonstrated smoking as a major contributor to the development and progression of CKD. For instance, a large retrospective study reported a fourfold increase in the risk of kidney failure among current smokers under 70 years, compared to never-smokers. Former smokers had a 3.3 times higher risk, with increased cumulative smoking exposure raising the risk and longer periods of smoking cessation reducing it [12]. Another study from a large Korean CKD cohort revealed that smokers with a cumulative exposure of 15–29 PYs and ≥30 PYs had a 1.48 and 1.94 times higher risk of CKD progression, respectively, compared to never-smokers. Moreover, longer smoking-free periods reduced adverse kidney outcomes, highlighting the importance of smoking cessation in CKD management [19]. Our findings align with these results but add an important nuance by emphasizing the additional risks posed by early smoking initiation and high cumulative exposure. However, unlike previous studies that focused solely on cumulative exposure, our analysis demonstrates that early smoking initiation independently exacerbates CKD risk, even among individuals with similar smoking intensities. Furthermore, the novel metric of age-standardized smoking intensity provides a refined framework for quantifying the compounded effects of prolonged exposure. These strengths position our study to fill a critical knowledge gap and inform targeted prevention strategies.
The biological mechanisms underlying the association between smoking and CKD are multifaceted. Smoking induces oxidative stress and inflammation, both key factors in CKD pathogenesis [20–23]. Nicotine and other toxic substances in tobacco can directly damage kidney cells, causing endothelial dysfunction and increased glomerular permeability [22]. This can result in proteinuria, a hallmark of kidney injury, and accelerate CKD progression. Smoking also promotes atherosclerosis, which can decrease blood flow and lead to ischemic injury in the kidneys. It has been shown to thicken arterial walls and speed up arteriosclerosis, reducing kidney perfusion and impairing function [24,25]. Additionally, smoking can alter the renin-angiotensin-aldosterone system (RAAS), a crucial pathway in blood pressure regulation and fluid balance [23,26]. Activation of RAAS by smoking can lead to hypertension, further contributing to kidney damage and disease progression. The combined effects of these mechanisms result in chronic inflammation and fibrosis, ultimately causing kidney function to decline.
In our study, individuals with higher cumulative smoking exposure exhibited a paradoxically higher baseline eGFR despite a greater prevalence of comorbidities such as hypertension, diabetes mellitus, and MetS. This finding may be attributed to reverse causation, wherein individuals with initially preserved kidney function had a longer duration of smoking exposure, leading to a higher PY accumulation. Since CKD progression is associated with decreased smoking rates due to declining health or medical interventions, individuals with lower eGFR may have already reduced or ceased smoking, resulting in a lower PY burden in that group. Moreover, the higher eGFR in heavy smokers may reflect glomerular hyperfiltration, a compensatory response to chronic vascular stressors such as smoking, which can ultimately contribute to kidney function decline [27,28]. This aligns with our findings that, despite a higher baseline eGFR, individuals with greater smoking exposure experienced worse long-term kidney outcomes, supporting the established link between chronic smoking, endothelial dysfunction, and CKD progression.
A large epidemiological study identified a significant inverse correlation between blood cotinine levels—a biomarker of nicotine exposure—and eGFR in adolescents who were actively smoking or exposed to secondhand smoke [29]. Adolescents exposed to cigarette smoke exhibited a lower average eGFR compared to their non-exposed counterparts, suggesting that the detrimental effects of smoking on kidney function may begin as early as adolescence. However, the precise mechanisms underlying adolescent smoking-related kidney impairment remain largely unexplored. Puberty is often regarded as a critical period with an increased risk of kidney function deterioration. Previous studies in pediatric CKD populations have demonstrated that eGFR decline accelerates significantly—up to tenfold—following the pubertal growth spurt, as estimated using creatinine-based equations [30]. A large multicenter cohort study of children with CKD further confirmed that the onset of puberty was associated with a more rapid decline in eGFR, a relationship that remained statistically significant even after adjusting for BMI [31]. The investigators proposed that rising sex hormone levels during puberty may directly contribute to kidney disease progression.
Hormonal changes during adolescence may also modulate the pathways through which smoking exerts nephrotoxic effects. Sex hormones regulate oxidative stress and inflammatory responses, two major contributors to smoking-induced kidney injury [32]. Nicotine exposure leads to increased production of reactive oxygen species (ROS) within kidney tissues and promotes inflammatory cell infiltration. Notably, testosterone amplifies ROS generation and stimulates the release of pro-inflammatory cytokines, such as tumor necrosis factor alpha, further exacerbating kidney damage [33]. Given this heightened vulnerability, earlier smoking initiation may subject individuals to prolonged oxidative stress, inflammation, and nicotine toxicity during a critical period of kidney development, increasing the long-term risk of chronic kidney injury. Of particular importance, our study population consisted predominantly of male adolescents, underscoring the potential for sex-specific vulnerability during puberty. Given the established role of testosterone in enhancing oxidative stress and inflammation, male adolescents who smoke or experience secondhand smoke exposure may be at an even greater risk of early kidney function decline.
From a clinical perspective, our findings underscore the importance of targeting smoking cessation efforts specifically at adolescents to mitigate the long-term risks of CKD and ESKD. Integrating questions about smoking initiation age into routine health assessments could aid clinicians in identifying high-risk individuals early, even before the onset of significant kidney impairment. Additionally, for CKD patients who began smoking early, tailored intervention strategies may help slow disease progression and improve long-term outcomes. Our data suggest that the risks associated with early smoking initiation are particularly pronounced in populations with a lower disease burden–such as individuals without MetS, diabetes mellitus, or hypertension. Targeted interventions in these generally healthy groups are crucial, as they may be less aware of the long-term health risks posed by smoking.
While our study offers valuable insights into the relationship between early smoking initiation and CKD progression, several limitations should be acknowledged. First, the observational design prevents us from establishing a direct causal relationship. Although we adjusted for multiple confounders, residual confounding remains a possibility. Second, the reliance on self-reported smoking data could introduce recall bias, especially in terms of age at initiation and PY calculations. Third, although a separate analysis of ex-smokers would have provided additional insights, we were unable to conduct this analysis due to data limitations. Future research incorporating ex-smoker-specific analyses would be valuable in further understanding smoking-related risks. Fourth, the definition of CKD in this study was based on a single eGFR measurement rather than the standard requirement of two consecutive measurements at least 3 months apart. While this does not fully align with the established diagnostic criteria, this approach was chosen to maximize the sample size and ensure sufficient statistical power. This may have led to the inclusion of individuals with transient reductions in eGFR rather than true CKD. Fifth, due to the restricted data access period under NHIS policy, we were unable to conduct additional post-hoc or interaction analyses that could have strengthened statistical inference. Lastly, the NHIS database lacks detailed information on prior cardiovascular diseases and medication use, as well as lifestyle and environmental factors that may influence CKD progression. Future research should incorporate these variables to provide a more comprehensive understanding of the effects of smoking on kidney health and to inform targeted prevention strategies. Despite these limitations, the large sample size, long-term follow-up, and focus on early-life exposure provide novel epidemiological insights into smoking-related kidney risk.
In conclusion, our study highlights the significant impact of age-standardized smoking intensity on ESKD risk, emphasizing the heightened vulnerability of individuals who begin smoking during adolescence. By prioritizing smoking prevention and cessation, we can reduce the long-term health burdens of smoking-related diseases and improve kidney health outcomes. Future studies should explore the mechanisms underlying these associations and consider the role of lifestyle and environmental factors in optimizing preventive strategies.
Supplementary Materials
Supplementary data are available at Kidney Research and Clinical Practice online (https://doi.org/10.23876/j.krcp.24.292).
Notes
Conflicts of interest
All authors have no conflicts of interest to declare.
Funding
The study was funded by grants from Seoul National University (3020210020). The funding sources played no role in the design or conduct of the study; collection, management, analysis, interpretation of the data; preparation, review, or approval of the manuscript.
Data sharing statement
The data are available from the National Health Insurance Service.
Authors’ contributions
Conceptualization: SGK, YK, KH
Data curation, Investigation: JHP, WYP, KJ, SH
Formal analysis: SP, KH, MK, EK, JHP, WYP, KJ, SH
Funding acquisition: DKK, YK
Methodology: JMC, SL, SC, HH
Project administration, Supervision: YK
Validation: DKK
Writing–original draft: SGK
Writing–review & editing: All authors
All authors read and approved the final manuscript.