Abstract
While observational studies have identified associations between gastroesophageal reflux disease (GERD) and laryngeal cancer (LC), the causal direction remains undetermined. This study employed a bidirectional 2-sample Mendelian randomization (MR) approach complemented by meta-analysis to investigate potential causal relationships between GERD and LC. Analysis leveraged publicly accessible genome-wide association study resources. Instrumental variables meeting MR assumptions were rigorously selected. The inverse variance weighted method served as primary analysis, with horizontal pleiotropy assessed through MR-Egger intercept tests. Sensitivity evaluations included Cochran Q test for heterogeneity and leave-one-out analysis. Findings from multiple databases were consolidated via meta-analysis. Inverse variance weighted estimates demonstrated no significant causal effect of GERD on LC risk (odds ratio [OR] = 1.11, 95% confidence interval [CI] = 0.83–1.48). In addition, reverse MR analysis revealed no causal impact of LC genetic liability on GERD development (odds ratio = 1.01, 95% CI = 0.99–1.02). All sensitivity analyses supported consistent results. This bidirectional MR investigation provides no genetic evidence for causal associations between GERD and LC in either direction.
Keywords: causal association, gastroesophageal reflux disease, laryngeal cancer, Mendelian randomization, meta-analysis
1. Introduction
Laryngeal cancer (LC) is a prevalent malignant neoplasm in the head and neck region, comprising approximately one-third of all head and neck cancers and ranking as the second most common respiratory system malignancy after lung cancer.[1,2] According to 2022 global cancer statistics, LC accounted for 188,960 new cases worldwide (0.9% of total cancer diagnoses), with 103,216 deaths (1.1% of cancer-related mortality).[3] The most well-established risk factors remain tobacco use and alcohol consumption, which exert a synergistic effect on carcinogenesis, while other etiological contributors are less conclusively established.[4] Given the substantial global health burden posed by LC, identification of modifiable risk factors is crucial for developing targeted prevention strategies.
Gastroesophageal reflux disease (GERD), one of the most prevalent gastrointestinal disorders, manifests as retrograde flow of gastric contents into the esophagus or supraesophageal regions (e.g., pharynx, larynx, and oral cavity), causing symptoms including acid regurgitation, heartburn, refractory cough, and asthma-like symptoms.[5] Although not life-threatening, GERD significantly impairs quality of life across disease stages, and its global prevalence and incidence have risen with enhanced socioeconomic conditions and changes in dietary habits and lifestyles.[6] Mounting clinical evidence establishes significant associations between GERD and malignant processes, particularly in oncogenesis and neoplastic advancement. Current epidemiological findings consistently identify heightened susceptibility to pulmonary, esophageal, oropharyngeal, and hypopharyngeal malignancies in populations with GERD.[7,8] Given the shared respiratory system between the lungs and larynx, as well as the anatomical proximity among pulmonary, esophageal, oral, pharyngeal, and laryngeal structures, epidemiological studies have investigated GERD–LC association, though with inconsistent results. For instance, multiple cohort and case-control studies indicate increased LC risk in GERD patients compared to non-GERD controls.[9,10] Additionally, anti-reflux procedures have been shown to reduce laryngeal squamous cell carcinoma (LSCC) risk in GERD populations,[11] and meta-analyses further demonstrate elevated LC risk in GERD patients compared to healthy controls.[12] However, David et al reported no increased LC risk in GERD patients.[13] Jerome et al believed that the association between GERD and LC had not been confirmed yet.[14] Critically, inherent limitations in traditional epidemiological investigations—including confounding variables and reverse causality—compromise their capacity to establish definitive causal relationships. Consequently, the etiological contribution of GERD to LC persists as an unresolved scientific inquiry.
Mendelian randomization (MR) methodology offers a powerful framework for causal inference between exposures and outcomes, overcoming inherent constraints of observational research designs.[15] This approach leverages genetic variants as instrumental variables (IVs), creating natural randomization comparable to randomized controlled trials, to strengthen epidemiological evidence. The proliferation of genome-wide association study (GWAS) resources has enabled broad implementation of 2-sample MR across disease domains. Our investigation applies GWAS-derived data within a bidirectional 2-sample MR design to elucidate GERD–LC causality. Given inter-database variability in instrumental variable selection, we implemented meta-analysis to synthesize multi-source effect estimates, strengthening inferential validity.[16]
2. Methods
This study was reported in accordance with the guidelines of the Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement.[17]
2.1. Study design
We applied bidirectional 2-sample MR to assess the causal relationship between GERD and LC, with the study design depicted in Figure 1. Using GERD as the exposure and LC as the outcome, we established a 2-sample MR framework. Single nucleotide polymorphisms (SNPs) significantly associated with GERD were selected as IVs from the FinnGen, UK Biobank, and GWAS catalog databases. Methodological validation included horizontal pleiotropy examination through MR-Egger intercept tests and heterogeneity quantification via Cochran Q statistic. “Leave-one-out” sensitivity testing validated result stability, preceding meta-analytic integration of cross-database MR findings. Subsequently, Exposure–outcome reversal analysis systematically designated LC as exposure and GERD as outcome. LC-associated SNPs were acquired from FinnGen and UK Biobank, repeating the established analytical workflow. Following methodological standards, this research employed publicly accessible GWAS summary statistics, exempt from additional ethical review as no individual-level data were analyzed.
Figure 1.
Study flow chart. (A) The 3 assumptions of MR analysis. (B) Flow chart of MR analysis with GERD as exposure and LC as outcome. (C) Flow chart of MR analysis with LC as exposure and GERD as outcome. GERD = gastroesophageal reflux disease.
The instrumental variable selection adhered to 3 MR foundational assumptions (Fig. 1A): Strong exposure–SNP associations; Independence from confounding factors; Outcome effects mediated solely via exposure pathways.[18]
2.2. Data sources
GERD genomic data were sourced from 3 databases: FinnGen,[19] UK Biobank,[20] and GWAS catalog.[21] The GWAS data of LC were obtained from FinnGen and UK Biobank.[20] All GWAS datasets were restricted to European populations to minimize population stratification bias. Comprehensive GWAS metadata are cataloged in Table 1.
Table 1.
Summary of the GWAS included in the study.
| Database | Number of case | Number of control | Sample size | Number of SNPs | Population | Year | Sources | |
|---|---|---|---|---|---|---|---|---|
| GERD | FinnGen | 36,025 | 423,785 | 459,810 | 20,462,956 | European | 2024 | https://storage.googleapis.com/finngen-public-data-r12/summary_stats/release/finngen_R12_K11_REFLUX.gz |
| UK Biobank | 14,372 | 441,976 | 456,348 | 11,831,932 | European | 2021 | PMID: 34737426 | |
| GWAS catalog | 129,080 | 473,524 | 602,604 | 2,324,711 | European | 2022 | https://www.ebi.ac.uk/gwas/studies/GCST90000514. | |
| LC | FinnGen | 374 | 259,583 | 259,957 | 18,706,739 | European | 2022 | https://storage.googleapis.com/finngen-public-data-r8/summary_stats/finngen_R8_C3_LARYNX_EXALLC.gz |
| UK Biobank | 341 | 455,935 | 456,276 | 11,831,924 | European | 2021 | PMID: 34737426 |
GERD = gastroesophageal reflux disease, GWAS = genome-wide association study, LC = laryngeal cancer, PMID = PubMed identifier, SNPs = single nucleotide polymorphisms.
2.3. Screening of IVs
Our MR analysis investigated potential causal associations between GERD and LC using genetic variants as IVs. The analysis followed 5 key steps:
To satisfy the first MR assumption (IVs strongly associate with exposure), SNPs reaching genome-wide significance (P < 5 × 10‐8) for the exposure were selected. When insufficient instruments were identified, SNPs with P < 5 × 10‐6 were included to ensure analytical robustness.[22]
To eliminate linkage disequilibrium bias, SNPs were clumped at r² < 0.001 within a 10,000 kb genomic window.[23]
Allele frequency harmonization was conducted to synchronize exposure–outcome datasets, excluding palindromic SNPs demonstrating intermediate allele frequencies with ambiguous strand orientation.
Per the second MR assumption, SNPs associated with known confounders (e.g., alcohol consumption, smoking) were excluded via the LDlink webtool (https://ldlink.nih.gov/?tab=ldtrait).[24]
To comply with the third MR assumption, SNPs showing direct associations with the outcome (P < .05) were removed.[25]
Instrument strength was assessed using the F-statistic[26]:
Here, β indicates the exposure-associated genetic effect magnitude with SE representing standard error. Genetic variants demonstrating F-statistics below 10 were systematically discarded to prevent weak instrument bias.
2.4. MR analysis
Analyses were conducted in R v4.4.2 with TwoSampleMR package implementation. Bidirectional 2-sample MR analyses implemented 5 methodologies: inverse variance weighting (IVW); weighted median; MR-Egger regression; simple mode; weighted mode. IVW served as the primary analytical approach, providing pooled causal estimates by combining SNP-specific effect estimates through meta-analysis.[27] This method achieves maximum reliability under the absence of horizontal pleiotropy.[28] MR-Egger intercept tests was utilized to identify horizontal pleiotropy.[29]
Heterogeneity evaluation employed Cochran Q statistic (P < .05 threshold). Sensitivity analyses through leave-one-out testing evaluated result stability by iteratively excluding individual SNPs and recalculating pooled effects. All analyses adopted α = 0.05 as the significance threshold. Direction-specific MR results from multiple databases were synthesized via meta-analysis using R v4.4.2 with the meta package.
3. Results
3.1. Effect of GERD on LC
Using GERD as the exposure and LC as the outcome, the IVW method was employed as the primary analytical approach. The comprehensive screening procedure for SNPs is provided in File S1 (Supplemental Digital Content, https://links.lww.com/MD/P906). MR analysis of SNPs derived from the FinnGen database showed no significant association between GERD and LC risk (FinnGen: odds ratio [OR] = 1.26, 95% confidence interval [CI] = 0.67–2.37, P = .47; UK Biobank: OR = 0.90, 95% CI = 0.46–1.79, P = .77). Analysis using SNPs from the UK Biobank database similarly indicated no association (FinnGen: OR = 1.05, 95% CI = 0.60–1.83, P = .86; UK Biobank: OR = 1.30, 95% CI = 0.69–2.46, P = .41). Evaluation of SNPs obtained from the GWAS catalog database also revealed no significant association (FinnGen: OR = 0.79, 95% CI = 0.27–2.34, P = .67; UK Biobank: OR = 1.44, 95% CI = 0.51–4.07, P = .49). The MR analyses across all databases demonstrated no significant causal link between GERD and LC development, as presented in Figures S1–S6 (Supplemental Digital Content, https://links.lww.com/MD/P905), with detailed results provided in Table 2.
Table 2.
MR analysis of the causality of GERD on LC.
| Exposure | Outcome | NSNPs | F-statistic | Method | OR (95% CI) | P-value |
|---|---|---|---|---|---|---|
| GERD (FinnGen database) | LC (FinnGen database) | 37 | 23.34 | MR-Egger | 1.33 (0.25, 6.93) | .74 |
| Weighted median | 1.07 (0.44, 2.62) | .87 | ||||
| IVW | 1.26 (0.67, 2.37) | .47 | ||||
| Simple mode | 0.69 (0.10, 4.67) | .70 | ||||
| Weighted mode | 0.69 (0.10, 4.89) | .71 | ||||
| GERD (UK Biobank database) | LC (UK Biobank database) | 34 | 23.48 | MR-Egger | 0.65 (0.13, 3.25) | .60 |
| Weighted median | 1.02 (0.41, 2.53) | .97 | ||||
| IVW | 0.90 (0.46, 1.79) | .77 | ||||
| Simple mode | 1.54 (0.28, 8.59) | .63 | ||||
| Weighted mode | 1.31 (0.27, 6.43) | .74 | ||||
| GERD (GWAS catalog database) | LC (FinnGen database) | 15 | 22.64 | MR-Egger | 1.18 (0.43, 3.20) | .76 |
| Weighted median | 0.98 (0.43, 2.23) | .96 | ||||
| IVW | 1.05 (0.60, 1.83) | .86 | ||||
| Simple mode | 1.12 (0.29, 4.42) | .87 | ||||
| Weighted mode | 0.96 (0.39, 2.38) | .93 | ||||
| GERD (FinnGen database) | LC (UK Biobank database) | 17 | 22.51 | MR-Egger | 1.11 (0.24, 5.09) | .89 |
| Weighted median | 1.18 (0.51, 2.73) | .70 | ||||
| IVW | 1.30 (0.69, 2.46) | .41 | ||||
| Simple mode | 1.13 (0.26, 4.93) | .87 | ||||
| Weighted mode | 1.13 (0.27, 4.73) | .86 | ||||
| GERD (UK Biobank database) | LC (FinnGen database) | 23 | 34.44 | MR-Egger | 0.01 (0.01, 64.62) | .28 |
| Weighted median | 0.82 (0.19, 3.57) | .79 | ||||
| IVW | 0.79 (0.27, 2.34) | .67 | ||||
| Simple mode | 1.92 (0.08, 46.15) | .70 | ||||
| Weighted mode | 0.32 (0.02, 6.41) | .46 | ||||
| GERD (GWAS catalog database) | LC (UK Biobank database) | 27 | 33.96 | MR-Egger | 0.17 (0.01, 165.66) | .62 |
| Weighted median | 1.70 (0.40, 7.17) | .47 | ||||
| IVW | 1.44 (0.51, 4.07) | .49 | ||||
| Simple mode | 2.36 (0.13, 41.82) | .56 | ||||
| Weighted mode | 1.83 (0.10, 32.34) | .68 |
CI = confidence interval, GERD = gastroesophageal reflux disease, GWAS = genome-wide association study, IVW = inverse variance weighted, LC = laryngeal cancer, MR = Mendelian randomization, NSNPs = number of single nucleotide polymorphisms, OR = odds ratio.
Since the I2 value was 0%, indicating the absence of significant heterogeneity, a fixed-effects model was utilized in the analysis. As shown in Figure 2, the meta-analysis resulted in an OR of 1.11, with a 95% CI spanning from 0.83 to 1.48. Consequently, we draw the conclusion that there is no causal connection between GERD and LC.
Figure 2.
Meta-analysis results of MR analyses from different database sources with GERD as exposure and LC as outcome. GERD = gastroesophageal reflux disease.
The outcomes presented in Table 3 show that there is no significant heterogeneity and horizontal pleiotropy among the SNPs within the 3 databases, which validates the reliability of the MR findings. When the MR analyses were conducted after the successive removal of SNPs, no notable alterations in the causal estimates were detected. Detailed results are provided in Figures S7–S12 (Supplemental Digital Content, https://links.lww.com/MD/P905).
Table 3.
Heterogeneity test and sensitivity analysis of the results of MR analysis.
| Exposure: GERD, outcome: LC | I and IV | I and V | II and IV | II and V | III and IV | III and V | |
|---|---|---|---|---|---|---|---|
| IVs | 37 | 34 | 15 | 17 | 23 | 27 | |
| Heterogeneity test (MR-Egger) | Cochran Q | 25.1451 | 17.0116 | 7.7024 | 8.4572 | 17.3478 | 17.4732 |
| Q_df | 35 | 32 | 13 | 15 | 21 | 25 | |
| P | .8905 | .9861 | .8624 | .9042 | .6898 | .8639 | |
| Heterogeneity test (IVW) | Cochran Q | 25.1492 | 17.2147 | 7.7727 | 8.5084 | 18.4916 | 17.8555 |
| Q_df | 36 | 33 | 14 | 16 | 22 | 26 | |
| P | .9123 | .9893 | .9008 | .9323 | .6765 | .8810 | |
| Horizontal pleiotropt test | Intercept | ‐0.0029 | 0.0204 | ‐0.0127 | 0.0140 | 0.1624 | 0.0685 |
| P | .9494 | .6553 | .7950 | .8240 | .2970 | .5419 | |
I = GERD of FinnGen database, II = GERD of UK Biobank database, III = GERD of GWAS catalog database, IV = LC of FinnGen database database, V = LC of UK Biobank database, GERD = gastroesophageal reflux disease, IVs = instrumental variables, IVW = inverse variance weighted, LC = laryngeal cancer, MR = Mendelian randomization.
3.2. Causal effects of LC on GERD
With LC designated as the exposure factor and GERD as the outcome variable, the IVW method was adopted as the principal analytical strategy. After harmonization with the GERD GWAS catalog database, both the FinnGen and UK Biobank databases for LC retained only 1 SNP each, which was insufficient for MR analysis. The detailed screening process of SNPs is described in File S2 (Supplemental Digital Content, https://links.lww.com/MD/P906). The MR analysis of SNPs derived from FinnGen revealed no significant association between LC and GERD risk (FinnGen: OR = 1.01, 95% CI = 0.98–1.04, P = .37; UK Biobank: OR = 1.02, 95% CI = 0.99–1.06, P = .20). Similarly, analysis using UK Biobank SNPs demonstrated no LC–GERD association (FinnGen: OR = 0.99, 95% CI = 0.98–1.01, P = .25; UK Biobank: OR = 1.02, 95% CI = 0.99–1.04, P = .15). The MR analyses across different databases uniformly demonstrated that there was no causal link between LC and GERD (Figures S13–S16, Supplemental Digital Content, https://links.lww.com/MD/P905), and the detailed results are presented in Table 4.
Table 4.
MR analysis of the causality of LC on GERD.
| Exposure | Outcome | NSNPs | F-statistic | Method | OR (95%CI) | P-value |
|---|---|---|---|---|---|---|
| LC (FinnGen database) | GERD (FinnGen database) | 4 | 22.87 | MR-Egger | 0.97 (0.87, 1.08) | .61 |
| Weighted median | 1.03 (0.99, 1.06) | .10 | ||||
| IVW | 1.01 (0.98, 1.04) | .37 | ||||
| Simple mode | 1.03 (0.97, 1.09) | .40 | ||||
| Weighted mode | 1.03 (0.99, 1.07) | .25 | ||||
| LC (FinnGen database) | GERD (UK Biobank database) | 5 | 23.13 | MR-Egger | 0.97 (0.90, 1.05) | .50 |
| Weighted median | 1.01 (0.97, 1.05) | .62 | ||||
| IVW | 1.02 (0.99, 1.06) | .20 | ||||
| Simple mode | 1.01 (0.95, 1.07) | .86 | ||||
| Weighted mode | 1.01 (0.96, 1.06) | .83 | ||||
| LC (UK Biobank database) | GERD (FinnGen database) | 8 | 22.20 | MR-Egger | 0.99 (0.96, 1.03) | .69 |
| Weighted median | 0.99 (0.97, 1.01) | .20 | ||||
| IVW | 0.99 (0.98, 1.01) | .25 | ||||
| Simple mode | 0.98 (0.96, 1.01) | .25 | ||||
| Weighted mode | 0.98 (0.96, 1.01) | .24 | ||||
| LC (UK Biobank database) | GERD (UK Biobank database) | 9 | 22.13 | MR-Egger | 1.02 (0.98, 1.07) | .31 |
| Weighted median | 1.01 (0.99, 1.04) | .32 | ||||
| IVW | 1.02 (0.99, 1.04) | .15 | ||||
| Simple mode | 1.01 (0.97, 1.05) | .67 | ||||
| Weighted mode | 1.01 (0.97, 1.05) | .68 |
CI = confidence interval, GERD = gastroesophageal reflux disease, IVW = inverse variance weighted, LC = laryngeal cancer, MR = Mendelian randomization, NSNPs = number of single nucleotide polymorphisms, OR = odds ratio.
With I2 = 48.1% indicating moderate heterogeneity, we applied a random-effects model for the analysis. The meta-analysis revealed an OR of 1.01 (95% CI = 0.99–1.02), as presented in Figure 3. We therefore conclude that no causal relationship exists between LC and GERD.
Figure 3.
Meta-analysis results of MR analyses from different database sources with LC as exposure and GERD as outcome. GERD = gastroesophageal reflux disease.
MR-Egger intercept tests were used to assess horizontal pleiotropy and Cochran Q to evaluate heterogeneity, revealing no significant evidence of either (Table 5). A leave-one-out sensitivity analysis was performed to evaluate the influence of individual SNPs from FinnGen and UK Biobank on causal estimates. Repeated MR analyses following sequential SNP removal demonstrated consistent causal estimates. Detailed results are provided in Figures S17–S20 (Supplemental Digital Content, https://links.lww.com/MD/P905).
Table 5.
Heterogeneity test and sensitivity analysis of the results of MR analysis.
| Exposure: LC, outcome: GRED | Ⅳ and Ⅰ | Ⅳ and Ⅱ | Ⅴ and Ⅰ | Ⅴ and Ⅱ | |
|---|---|---|---|---|---|
| IVs | 4 | 5 | 8 | 9 | |
| Heterogeneity test (MR-Egger) | Cochran Q | 2.8198 | 2.9285 | 3.1520 | 4.6028 |
| Q_df | 2 | 3 | 6 | 7 | |
| P | .2442 | .4028 | .7895 | 7083 | |
| Heterogeneity test (IVW) | Cochran Q | 3.9216 | 5.1394 | 3.1605 | 4.8041 |
| Q_df | 3 | 4 | 7 | 8 | |
| P | .2701 | .2733 | .8698 | .7783 | |
| Horizontal pleiotropt test | Intercept | 0.0275 | 0.0393 | ‐0.0010 | ‐0.0058 |
| P | .4699 | .2337 | .9295 | .6672 | |
I = GERD of FinnGen database, II = GERD of UK Biobank database, IV = LC of FinnGen database database, V = LC of UK Biobank database, GERD = gastroesophageal reflux disease, IVs = instrumental variables, IVW = inverse variance weighted, LC = laryngeal cancer, MR = Mendelian randomization.
4. Discussion
In this research, using of publicly accessible GWAS summary data, the causal connection between GERD and the risk of LC was explored via bidirectional 2-sample MR in combination with meta-analysis. The results revealed no significant association between GERD and LC incidence, with reverse analyses also demonstrating no causal relationship. Sensitivity analyses verified the reliability of these findings.
Over recent years, numerous epidemiological investigations have pointed out the correlation between GERD and LC, which is inconsistent with our research results. The U.S. Department of Veterans Affairs conducted a retrospective matched case-control study among U.S. veterans. This study included 8228 hospitalized LC patients compared with 32,912 hospitalized controls, and 9292 outpatient LC patients compared with 37,168 cancer-free outpatient controls. The conclusion was that among U.S. veterans, hospitalized patients who had GERD faced an elevated risk of LC (OR = 2.40, 95% CI = 2.15–2.69), and outpatient patients with GERD also had an increased likelihood of developing LC (OR = 2.31, 95% CI = 2.10–2.53).[30] In addition, Sean et al discovered that patients suffering from GERD had a notably higher risk of LC or carcinoma in situ of the larynx (OR = 1.33, 95% CI = 1.04–1.59).[31] Even though these studies made matching for age, gender, race, smoking habits, and alcohol consumption levels in an attempt to minimize the confounding effects, it was still challenging to evade the inherent drawbacks of retrospective studies. Consequently, another study from the National Institutes of Health-American Association of Retired Persons Diet and Health Study, which altogether included 116,476 participants with GERD, evaluated the relationship between GERD and subsequent LSCC from a prospective viewpoint. The findings demonstrated that during the 16-year follow-up period, there was a positive link between GERD and the risk of newly diagnosed LSCC (HR = 1.91, 95% CI = 1.24–2.94).[32] Moreover, a meta-analysis summarized 15 case-control studies and found that individuals with GERD had a 2.37-fold increased risk of laryngeal malignancies (OR = 2.37, 95% CI = 1.79–3.14).[12] The above real-world data indicate that patients with GERD may have a higher susceptibility to LC. However, the results of a retrospective matched case-control study from northeastern Ohio in the United States showed that when GERD was included based on the International Classification of Diseases-9 codes diagnostic criteria, the risk of LC in patients with GERD was slightly reduced (OR = 0.52, 95% CI = 0.24–1.09).[33]
Nevertheless, not every observational study has been able to prove a direct causal link between GERD and LC. Magnus et al carried out a large-scale retrospective cohort study by utilizing data from the Swedish inpatient registry. This particular cohort consisted of 66,965 patients who had discharge diagnoses of heartburn, diaphragmatic hernia, or esophagitis. Once cohort members diagnosed with alcoholism were excluded, there was no notable rise in the risk of LC when compared to the general population.[34] A population-based prospective case-control study from the Carolina Head and Neck Cancer Epidemiology study found that there was no association between GERD and LC among study subjects who were neither heavy smokers nor heavy drinkers.[35]
Considering that association studies cannot answer the question of causal relationships, it is difficult to confirm the causal relationship between GERD and LC based solely on observational studies. In conclusion, the above findings should be interpreted with caution. Contrary to the results of the majority of observational studies, our research did not identify a causal relationship between GERD and LC. Obviously, different analytical methods may lead to differences in findings between these studies and ours. Observational studies are especially prone to the influence of confounding factors, whereas MR analysis has the ability to circumvent this shortcoming, thereby enabling the generation of relatively precise causal determinations.
In observational studies, there may be several reasons for the association between GERD and LC. First, the direct damage caused by the reflux might have a certain influence on the development of LC. Long-term GERD can cause damage to the esophageal mucosa, resulting in esophagitis, metaplasia (Barrett esophagus), and eventually esophageal adenocarcinoma.[36,37] Compared with the esophagus, the squamous epithelial cells of the larynx are more vulnerable to the damage of duodenogastric contents.[38] However, the occurrence of GERD-related esophageal cancer hinges on the progression of Barrett esophagus as an intermediate stage prior to the onset of adenocarcinoma, and it has no connection with esophageal squamous cell carcinoma.[39] Since LC is mainly laryngeal squamous cell carcinoma, further research is still needed to confirm whether the direct damage caused by the reflux is related to LC. Second, the chronic inflammatory effect caused by the reflux may be related to the occurrence of LC.[40] Duodenogastric contents include hydrochloric acid, pepsin, and bile salts. Some studies have shown that the presence of pepsin acts as a continuous or low-intensity stimulus to the laryngeal mucosa. This long-term and repetitive chemical stimulation leads to damage of the laryngeal epithelium and structural alterations, making the laryngeal mucosa in a state of chronic inflammation, leading to its tumorigenic transformation, increasing the risk of LC, and possibly playing an important role in the malignant transformation from precancerous lesions to LC.[41,42] Meanwhile, for some GERD patients with Helicobacter pylori (H pylori) infection, GERD can carry H pylori colonized in the gastric mucosa to the larynx.[43] GERD can also provide an acidic environment for the larynx, allowing H pylori to survive in the larynx.[44] Some studies have shown that H pylori may trigger systemic inflammation and promote tumorigenesis, which may have a significant influence on the development of LC.[45]
Our MR study possesses several notable advantages. First, as far as we are aware, this represents the initial 2-sample MR analysis grounded in extensive large-scale GWAS data aimed at assessing the causal relationship between GERD and LC. Compared with earlier observational studies, the MR analysis is capable of effectively reducing potential biases. These biases encompass confounding factors and reverse causality, thus enhancing the causal inference that can be drawn. Second, the GWAS datasets for GERD and LC utilized in this study predominantly originate from European-ancestry populations. This characteristic helps to minimize the influence of population stratification, which could otherwise confound the results. Third, a meta-analysis was adopted to summarize the results of MR analyses from various databases. By doing so, it renders the conclusions more reliable and robust. In addition, multiple sensitivity analyses were carried out to validate the robustness and consistency of the findings. However, we also recognize certain limitations of our study. First, this investigation only encompasses a single population. Consequently, the representativeness of the results requires further verification across the entire population. Second, although we did not uncover a causal relationship between GERD and LC, some studies have clinically observed a correlation between them. This may be due to the differences in the diagnostic criteria of GERD in the population statistics involved in the GWAS database used and the previous partial observational studies, as well as the failure to clearly distinguish between gastroesophageal reflux and laryngopharyngeal reflux. Third, a larger sample size and more advanced methods are necessary to confirm the results and fully demonstrate the statistical power. Finally, while GWAS can offer new perspectives on the genes related to LC, the specific mechanisms still demand further investigation.
In conclusion, our MR analysis shows that genetically-predicted GERD has no causal effect on LC, which contradicts the outcomes of the majority of observational studies. To confirm the accuracy of our findings, future studies need to be conducted based on higher-quality genome-wide association study data and more advanced methods.
Acknowledgments
We gratefully recognize the Department of Otolaryngology Head and Neck Surgery of Beijing Anzhen Hospital of Capital Medical University & Nanchong Central Hospital.
Author contributions
Conceptualization: Xi Zhang, Min Yan.
Data curation: Xi Zhang, Min Yan.
Formal analysis: Xi Zhang, Min Yan.
Funding acquisition: Xi Zhang.
Investigation: Xi Zhang.
Methodology: Xi Zhang, Min Yan.
Project administration: Xi Zhang.
Resources: Xi Zhang.
Software: Xi Zhang.
Supervision: Min Yan.
Validation: Min Yan.
Visualization: Min Yan.
Writing – original draft: Xi Zhang.
Writing – review & editing: Xi Zhang, Min Yan.
Supplementary Material
Abbreviations:
- CI
- confidence interval
- GERD
- gastroesophageal reflux disease
- GWAS
- genome-wide association study
- IVs
- instrumental variables
- IVW
- inverse variance weighted
- LC
- laryngeal cancer
- LD
- linkage disequilibrium
- LSCC
- laryngeal squamous cell carcinoma
- MR
- Mendelian randomization
- OR
- odds ratio
- SNPs
- single nucleotide polymorphisms
The authors have no funding and conflicts of interest to disclose.
All data generated or analyzed during this study are included in this published article [and its supplementary information files].
Supplemental Digital Content is available for this article.
How to cite this article: Zhang X, Yan M. Gastroesophageal reflux disease and risk of laryngeal cancer: A Mendelian randomization study. Medicine 2025;104:36(e44395).
Min Yan accepts full responsibility for the conduct of the study, had access to the data, and controlled the decision to publish.
References
- [1].Piotrowski I, Zhu X, Saccon TD, et al. miRNAs as biomarkers for diagnosing and predicting survival of head and neck squamous cell carcinoma patients. Cancers (Basel). 2021;13:3980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Megwalu UC, Sikora AG. Survival outcomes in advanced laryngeal cancer. JAMA Otolaryngol Head Neck Surg. 2014;140:855–60. [DOI] [PubMed] [Google Scholar]
- [3].Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74:229–63. [DOI] [PubMed] [Google Scholar]
- [4].Steuer CE, El-Deiry M, Parks JR, Higgins KA, Saba NF. An update on larynx cancer. CA Cancer J Clin. 2017;67:31–50. [DOI] [PubMed] [Google Scholar]
- [5].Zheng Z, Shang Y, Wang N, et al. Current advancement on the dynamic mechanism of gastroesophageal reflux disease. Int J Biol Sci. 2021;17:4154–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Eusebi LH, Ratnakumaran R, Yuan Y, Solaymani-Dodaran M, Bazzoli F, Ford AC. Global prevalence of, and risk factors for, gastro-oesophageal reflux symptoms: a meta-analysis. Gut. 2018;67:430–40. [DOI] [PubMed] [Google Scholar]
- [7].Liu Y, Lai H, Zhang R, Xia L, Liu L. Causal relationship between gastro-esophageal reflux disease and risk of lung cancer: insights from multivariable Mendelian randomization and mediation analysis. Int J Epidemiol. 2023;52:1435–47. [DOI] [PubMed] [Google Scholar]
- [8].Wu G, Liu Y, Ning D, et al. Unraveling the causality between gastroesophageal reflux disease and increased cancer risk: evidence from the UK Biobank and GWAS consortia. BMC Med. 2024;22:323. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Tran CL, Han M, Kim B, Park EY, Kim YI, Oh JK. Gastroesophageal reflux disease and risk of cancer: findings from the Korean National Health Screening Cohort. Cancer Med. 2023;12:19163–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Riley CA, Wu EL, Hsieh MC, Marino MJ, Wu XC, McCoul ED. Association of gastroesophageal reflux with malignancy of the upper aerodigestive tract in elderly patients. JAMA Otolaryngol Head Neck Surg. 2018;144:140–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Yanes M, Santoni G, Maret-Ouda J, et al. Laryngeal and pharyngeal squamous cell carcinoma after antireflux surgery in the 5 Nordic countries. Ann Surg. 2022;276:e79–85. [DOI] [PubMed] [Google Scholar]
- [12].Parsel SM, Wu EL, Riley CA, McCoul ED. Gastroesophageal and laryngopharyngeal reflux associated with laryngeal malignancy: a systematic review and meta-analysis. Clin Gastroenterol Hepatol. 2019;17:1253–64.e5. [DOI] [PubMed] [Google Scholar]
- [13].Francis DO, Maynard C, Weymuller EA, Reiber G, Merati AL, Yueh B. Reevaluation of gastroesophageal reflux disease as a risk factor for laryngeal cancer. Laryngoscope. 2011;121:102–5. [DOI] [PubMed] [Google Scholar]
- [14].Lechien JR, Eun YG, Hans S, Saussez S. The association between reflux and laryngeal neoplasia is still not demonstrated. Otolaryngol Head Neck Surg. 2020;163:1284–5. [DOI] [PubMed] [Google Scholar]
- [15].Emdin CA, Khera AV, Kathiresan S. Mendelian randomization. JAMA. 2017;318:1925–6. [DOI] [PubMed] [Google Scholar]
- [16].Luo J, Xu Z, Noordam R, van Heemst D, Li-Gao R. Depression and inflammatory bowel disease: a bidirectional two-sample Mendelian randomization study. J Crohns Colitis. 2022;16:633–42. [DOI] [PubMed] [Google Scholar]
- [17].Skrivankova VW, Richmond RC, Woolf BAR, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: the STROBE-MR statement. JAMA. 2021;326:1614–21. [DOI] [PubMed] [Google Scholar]
- [18].Haycock PC, Burgess S, Wade KH, Bowden J, Relton C, Davey Smith G. Best (but oft-forgotten) practices: the design, analysis, and interpretation of Mendelian randomization studies. Am J Clin Nutr. 2016;103:965–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Kurki MI, Karjalainen J, Palta P, et al. FinnGen provides genetic insights from a well-phenotyped isolated population [published correction appears in Nature. March 2023;615(7952):E19. doi:10.1038/s41586-023-05837-8]. Nature. 2023;613:508–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Jiang L, Zheng Z, Fang H, Yang J. A generalized linear mixed model association tool for biobank-scale data. Nat Genet. 2021;53:1616–21. [DOI] [PubMed] [Google Scholar]
- [21].Ong JS, An J, Han X, et al. Multitrait genetic association analysis identifies 50 new risk loci for gastro-oesophageal reflux, seven new loci for Barrett’s oesophagus and provides insights into clinical heterogeneity in reflux diagnosis. Gut. 2022;71:1053–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Zhao ZN, Zhang ZQ, Wang QQ, et al. Genetic predisposition to periodontitis and risk of migraine: a two-sample Mendelian randomization study. Neurol Ther. 2023;12:1159–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Chen J, Chen R, Xiang S, et al. Cigarette smoking and schizophrenia: Mendelian randomisation study. Br J Psychiatry. 2021;218:98–103. [DOI] [PubMed] [Google Scholar]
- [24].Lin SH, Brown DW, Machiela MJ. LDtrait: an online tool for identifying published phenotype associations in linkage disequilibrium. Cancer Res. 2020;80:3443–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Shi Z, Li Z, Wang K, Yang F. The causal role of gastroesophageal reflux disease in endometriosis: a bidirectional Mendelian randomization study. Front Med (Lausanne). 2024;11:1440157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Qian J, Xu H, Liu J, Zheng Y. Associations of cholecystectomy with the risk of gastroesophageal reflux disease: a Mendelian randomization study. Int J Surg. 2024;110:6836–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Magnus MC, Miliku K, Bauer A, et al. Vitamin D and risk of pregnancy related hypertensive disorders: mendelian randomisation study. BMJ. 2018;361:k2167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Long Y, Tang L, Zhou Y, Zhao S, Zhu H. Causal relationship between gut microbiota and cancers: a two-sample Mendelian randomisation study. BMC Med. 2023;21:66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Higbee DH, Granell R, Sanderson E, Davey Smith G, Dodd JW. Lung function and cardiovascular disease: a two-sample Mendelian randomisation study. Eur Respir J. 2021;58:2003196. [DOI] [PubMed] [Google Scholar]
- [30].El-Serag HB, Hepworth EJ, Lee P, Sonnenberg A. Gastroesophageal reflux disease is a risk factor for laryngeal and pharyngeal cancer. Am J Gastroenterol. 2001;96:2013–8. [DOI] [PubMed] [Google Scholar]
- [31].Parsel SM, Iarocci AL, Gastañaduy M, Winters RD, Marino JP, McCoul ED. Reflux disease and laryngeal neoplasia in nonsmokers and nondrinkers. Otolaryngol Head Neck Surg. 2020;163:560–2. [DOI] [PubMed] [Google Scholar]
- [32].Wang SM, Freedman ND, Katki HA, et al. Gastroesophageal reflux disease: a risk factor for laryngeal squamous cell carcinoma and esophageal squamous cell carcinoma in the NIH-AARP Diet and Health Study cohort. Cancer. 2021;127:1871–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Vaezi MF, Qadeer MA, Lopez R, Colabianchi N. Laryngeal cancer and gastroesophageal reflux disease: a case-control study. Am J Med. 2006;119:768–76. [DOI] [PubMed] [Google Scholar]
- [34].Nilsson M, Chow WH, Lindblad M, Ye W. No association between gastroesophageal reflux and cancers of the larynx and pharynx. Cancer Epidemiol Biomarkers Prev. 2005;14:1194–7. [DOI] [PubMed] [Google Scholar]
- [35].Busch EL, Zevallos JP, Olshan AF. Gastroesophageal reflux disease and odds of head and neck squamous cell carcinoma in North Carolina. Laryngoscope. 2016;126:1091–6. [DOI] [PubMed] [Google Scholar]
- [36].Vakil N, van Zanten SV, Kahrilas P, Dent J, Jones R. The Montreal definition and classification of gastroesophageal reflux disease: a global evidence-based consensus. Am J Gastroenterol. 2006;101:1900–20; quiz 1943. [DOI] [PubMed] [Google Scholar]
- [37].Ness-Jensen E, Lagergren J. Tobacco smoking, alcohol consumption and gastro-oesophageal reflux disease. Best Pract Res Clin Gastroenterol. 2017;31:501–8. [DOI] [PubMed] [Google Scholar]
- [38].Asaoka D, Nagahara A, Matsumoto K, Hojo M, Watanabe S. Current perspectives on reflux laryngitis. Clin J Gastroenterol. 2014;7:471–5. [DOI] [PubMed] [Google Scholar]
- [39].Lagergren J, Smyth E, Cunningham D, Lagergren P. Oesophageal cancer. Lancet. 2017;390:2383–96. [DOI] [PubMed] [Google Scholar]
- [40].Assimakopoulos D, Patrikakos G. The role of gastroesophageal reflux in the pathogenesis of laryngeal carcinoma. Am J Otolaryngol. 2002;23:351–7. [DOI] [PubMed] [Google Scholar]
- [41].Li H, Zhang S, Zhou S, et al. Pepsin enhances glycolysis to promote malignant transformation of vocal fold leukoplakia epithelial cells with dysplasia. Eur Arch Otorhinolaryngol. 2023;280:1841–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Yin CY, Zhang SS, Zhong JT, Zhou SH. Pepsin and laryngeal and hypopharyngeal carcinomas. Clin Exp Otorhinolaryngol. 2021;14:159–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [43].Genç R, Çağli S, Yüce I, et al. The role of H. pylori in the development of laryngeal squamous cell carcinoma. Dis Markers. 2013;35:447–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Huang Y, Gu M, Wu Q, et al. Is laryngeal squamous cell carcinoma related to Helicobacter pylori? Front Oncol. 2022;12:790997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [45].Chen M, Fang Y, Cheng L, Wu H. Helicobacter pylori is associated with poor prognosis of laryngeal precancerous lesion. Auris Nasus Larynx. 2020;47:268–75. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.