Lung cancer is the second most commonly diagnosed cancer worldwide and the leading cause of cancer mortality, accounting for 18% of deaths in both men and women (1, 2). Chronic obstructive pulmonary disease (COPD) increases the risk of lung cancer due to shared etiologies, such as smoking, environmental pollution exposure, genetic predisposition, and inflammation pathways (3–5). Studies have also shown that COPD is an independent risk factor for lung cancer, with a six-fold higher risk regardless of smoking status (6, 7). However, COPD is a heterogeneous disease with various clinical phenotypes and comorbidities that can worsen its clinical outcomes and cause systemic inflammation (8). Type 2 diabetes mellitus (T2DM) is one of the most common comorbidities in patients with COPD. A cohort study of 2,164 patients with COPD found that diabetes mellitus (DM) increased mortality (9).

Diabetes mellitus (DM) is a systemic disorder characterized by hyperglycemia, inflammation, and oxidative stress that can affect organs such as the kidney, retina, and vascular system (10–12). DM may also influence lung cancer through mechanisms, such as hyperinsulinemia, hyperglycemia, or chronic inflammation, which have been associated with cancer progression (13, 14). However, the association between DM and lung cancer is inconclusive. While some studies did not find an increased risk of lung cancer in patients with DM (15, 16), other studies suggested a higher risk (17–19). Since T2DM is prevalent in patients with COPD and may increase the risk of developing lung cancer, this possible association requires further investigation.

To determine whether T2DM is associated with an increased risk of lung cancer in patients with COPD, we conducted a retrospective analysis on two cohorts: the National Health Insurance Service-National Sample Cohort (NHIS-NSC) of Korea and the Common Data Model (CDM) database of a university hospital.

Based on the NHIS-NSC database, we enrolled patients who were newly diagnosed with COPD and aged between 35 and 90 years from 2002 to 2018 (n = 13,939). We excluded patients who were diagnosed with COPD during a washout period from 2002 to 2003 (n = 2,184), those with missing values for smoking status or BMI (n = 1,999), and those who were diagnosed with lung cancer before or within one month after a COPD diagnosis (n = 299). Each patient was followed up until the development of lung cancer, death, or the end of 2019, whichever occurred first. Among the remaining patients, 579 with COPD were diagnosed with lung cancer in the NIHS-NSC database. We performed 1:5 propensity score matching to select a control group without lung cancer patients among eligible patients with COPD. In total, 3,474 patients were selected, consisting of 1,922 without T2DM and 1,552 with T2DM (Figure 1A).

From the CDM cohort, there were 6,057 patients, aged 35–90 years, who were newly diagnosed with COPD between 2001 and 2018. We excluded patients diagnosed with COPD during a washout period from 2001 to 2003 (n = 888) and those diagnosed with lung cancer before or within one month after a COPD diagnosis (n = 267). Each patient was followed up until the development of lung cancer, death, or the end of 2019, whichever occurred first. Finally, there were 143 patients with COPD diagnosed with lung cancer in the CDM cohort. We performed 1:5 propensity score matching to select a control group without lung cancer patients among eligible patients with COPD, using the same criteria as in the NHIS dataset. In total, 858 patients were selected, including 648 without T2DM and 210 with T2DM (Figure 1B).

Table 1 summarizes the baseline demographic characteristics of patients with COPD with and without T2DM in each of the two cohorts. In the NHIS-NSC, patients with COPD and T2DM had a higher mean BMI and a higher proportion of rural residents than those without T2DM. There was no significant difference in age, sex, smoking status, or smoking pack-years. In the CDM cohort, patients with COPD and T2DM had a higher mean BMI and white blood cell counts than those without T2DM. There were more cases of CCI values ≥1 in patients with COPD and T2DM. There was no significant difference in age, sex, or predicted FEV₁, FVC, or FEV₁/FVC ratio. Smoking data could not be compared due to the absence of information in the CDM cohort.

Results of the Kaplan–Meier analysis showed a higher cumulative lung cancer incidence in patients with COPD and T2DM than in those without T2DM (NHIS-NSC: P = 0.014; CDM: P = 0.032) (Figure 2).

In the NHIS-NSC, T2DM was associated with an increased risk of lung cancer (HR, 1.23; 95% CI, 1.04–1.44). When adjusting for age, sex, smoking pack-years, household income, residential area, BMI, and CCI, the risk of lung cancer was significantly higher in patients with COPD and T2DM than in those without T2DM (adjusted HR [aHR], 1.20; 95% CI, 1.02–1.41; Figure 3A). Among patients with COPD from the CDM cohort, T2DM was independently associated with an increased risk of lung cancer after adjusting for age, sex, BMI, and CCI (aHR, 1.45; 95% CI, 1.02–2.07; Figure 3B).

A subgroup analysis was performed in patients with COPD and T2DM to determine whether there was a difference in the risk of lung cancer according to age, sex, BMI, CCI, smoking pack-years, household income, and residential area (Table 2). In the NHIS-NSC, the risk of lung cancer was higher in current smokers than in never-smokers (aHR, 1.45; 95% CI, 1.09–1.91); in smokers with ≥30 pack-years than in never-smokers (aHR, 1.82; 95% CI, 1.49–2.25); and in rural residents than in metropolitan residents (aHR, 1.33; 95% CI, 1.06–1.68).

In this study, through two independent cohort analyses, we found that patients with COPD and T2DM had a higher risk of developing lung cancer than those without T2DM, even after adjusting for possible confounding factors. Based on a subgroup analysis of patients with COPD and T2DM, the risk of lung cancer was increased among smokers, which is commonly observed in lung cancer prevalence (1). Additionally, in patients with COPD and T2DM, the risk of lung cancer was higher in rural residents than in metropolitan residents. These findings are consistent with those of a study by Jenkins et al. that reported increased lung cancer incidence and mortality rates in rural areas (21). In rural areas, people are less likely to consult a doctor due to low accessibility to hospitals, and it is difficult to participate in smoking cessation and health programs, which can increase the risk of cancer diagnosis (22, 23).

The hypothesis for our study was that T2DM, a well-known comorbidity of COPD, not only impacts the clinical outcomes of COPD but also increases the risk of lung cancer in patients with COPD. Various studies have investigated the possibility that diabetic microangiopathy can affect the lungs, leading to histological changes, functional abnormalities, and chronic inflammation at the molecular level (10, 24, 25). Recent case-control studies have also demonstrated that T2DM significantly decreases pulmonary function compared to healthy controls (26, 27). An experimental model has demonstrated that hyperglycemia can increase levels of inflammatory cytokines, including interleukin-6 and tumor necrosis factor-alpha, which play a crucial role in epithelial-to-mesenchymal transition and tumorigenesis (28). In our CDM cohort, patients with COPD and T2DM had higher mean white blood cell counts and, although not statistically significant, lower mean hemoglobin and higher mean C-reactive protein levels than those in patients without T2DM. These findings may indicate that T2DM increases systemic inflammation in patients with COPD.

Several retrospective cohort studies using the National Health Insurance (NHI) database in Taiwan have reported associations between diabetes, lung cancer, and COPD. Two studies, also using the NHI database, reported that diabetes increased the risk of lung cancer (18, 19). Tseng reported that age, sex, and COPD were positively associated with lung cancer among patients with DM (19). However, Shen et al. concluded that T2DM could have a protective effect against lung cancer in patients with COPD (29), which contradicts our findings. They speculated that the anti-inflammatory effect of anti-diabetic drugs might prevent lung cancer. Although many studies have reported that metformin, one of the most widely prescribed anti-diabetic drugs, reduces the risk of cancer (30, 31), a large observational study in the United Kingdom showed that metformin use was not associated with a decreased risk of lung cancer (32). Additionally, from our perspective, the chronic inflammation in T2DM, along with the chronic inflammation in COPD, is more likely to increase the risk of lung cancer.

Another important point is that while Shen et al. analyzed only the NHI data of Taiwan, we obtained the same results by analyzing the NHIS-NSC of Korea as well as the CDM cohort data of a university hospital. These two cohorts had complementary advantages. Although the NHIS-NSC is representative of the country, it has the disadvantage that the diagnosis of each disease requires manipulation of the data. In comparison, since the CDM cohort included lung function data and diagnosis by a relatively small number of respiratory specialists, the accuracy of the COPD diagnosis was high. While smoking history, the most important risk factor for lung cancer, is not available in the NHI database of Taiwan, the NHIS-NSC data included smoking history obtained during national health checkups.

Type 2 diabetes mellitus is a common comorbidity in patients with COPD and may increase the risk of lung cancer. COPD can activate inflammation pathways that increase oxidative stress, leading to DNA damage, which is a known trigger for lung cancer (33). Some studies have suggested that COPD may also induce T2DM, increase the risk for metabolic syndrome, and cause insulin resistance (24, 34). Although the causes and effects of the association between DM and COPD are unclear, they may influence the development of lung cancer.

Tobacco is a potent carcinogen, and its harmful effects, when smoked or exposed to nearby individuals intermittently, are well-established. In contrast, although the impact of T2DM on tumorigenesis may be less significant, persistent and frequent exposure to hyperglycemia can contribute to the development of cancer (35). Our study found that, after adjusting for smoking, a potential confounding factor in clinical settings, the HR of DM on COPD remained significant. Furthermore, results our subgroup analysis of patients with COPD and T2DM showed that current smokers with T2DM had a higher risk of lung cancer, suggesting a potential additive effect on the development of cancer. Finally, the higher risk of lung cancer in rural areas may be not only due to poor participation in smoking cessation programs but also inevitable glucose control for patients with T2DM due to limited access to medical resources.

This study has several strengths. First, the accuracy of the study was improved by using two independent cohorts. The NHIS-NSC dataset is a large, representative data source based on the national population, and it includes risk factors for lung cancer such as smoking status, household income, residential area, and BMI. In the CDM cohort, a spirometry-based COPD definition was additionally applied, and detailed medication information based on hospital data was available when selecting study patients. Diagnoses and prescriptions were confirmed by pulmonologists and endocrinologists, ensuring the high reliability of the diagnosis. The long follow-up period is another strength of this study.

However, our study has some limitations. In the CDM data, there is a lack of information about smoking, which is a major risk factor for lung cancer. Smoking data were recorded as text files in the electronic medical forms and could not be extracted. Nevertheless, similar to other studies representative of Korea (36), 92% of patients with COPD in this study were men, and given the high smoking rate among men in Korea, it is very likely that most were current or former smokers. Moreover, the results were very similar to those of the NHIS-NSC, which was analyzed using smoking status and pack-years. Second, although this study investigated a large dataset of patients with COPD, caution is needed in generalizing its results due to the limitations of a retrospective epidemiologic study. Third, the type of treatment for COPD and T2DM and its administration period, the degree of symptom and glycemic control, and the duration of T2DM may affect lung cancer incidence. However, analysis of these factors was not possible based on the current dataset.

Patients with COPD and T2DM have an increased risk of developing lung cancer compared with those without T2DM. Furthermore, among patients with both COPD and T2DM, the risk of lung cancer was found to be higher in current smokers and in smokers with ≥30 pack-years than in never-smokers, and in rural residents than in metropolitan residents. Although further studies are necessary, it is prudent to carefully consider the possibility of lung cancer during the follow-up of patients with COPD and T2DM.

The NHIS-NSC and CDM data may be provided through a remote virtualization room when an application is made to the National Health Insurance Service (https://nhiss.nhis.or.kr/bd/ab/bdaba001cv.do) and the Ewha Womans University Medical Center (http://part.eumc.ac.kr/dept/RDPC/index.do), respectively, and is approved. However, the use of the data is only available in Republic of Korea.

The studies involving human participants were reviewed and approved by the Institutional review board (IRB) of Ewha Womans University Medical Center, Seoul, Republic of Korea (IRB number: SEUMC2021-11-007) and the IRB waived the need to obtain informed consent considering the retrospective nature of the study. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

JL and JYL were the guarantors, took responsibility for the content of the manuscript, including the data and analysis, and were responsible for the acquisition of data and study supervision. EHK and JYL analyzed the data and created the figures. NK and EHK wrote the first draft of the manuscript. All authors contributed substantially to the study concept and design, interpreted the analysis results, critically revised the manuscript, provided supervision for the final manuscript, and approved the final draft of the manuscript.