Family History of Esophageal Cancer Modifies the Association of Serum Lipids and Malignant Esophageal Lesions: A Nested Case-Control Study from the “Endoscopic Screening for Esophageal Cancer in China” Trial
Introduction
Esophageal squamous cell carcinoma (ESCC) accounts for 90% of esophageal cancer (EC) cases in China, posing a significant public health burden. While epidemiological studies have long explored associations between lipid metabolism and carcinogenesis, results remain inconsistent across cancer types, populations, and lipid components. Prior research on serum lipids and ESCC risk is limited, with one Korean cohort study reporting weak inverse associations in men but no significant links in women. This study addresses critical gaps by evaluating how genetic susceptibility, proxied by family history of EC, modifies the relationship between serum lipid profiles and malignant esophageal lesions within a high-risk Chinese population.
The “Endoscopic Screening for Esophageal Cancer in China” (ESECC) trial provided a unique platform for this nested case-control analysis. Conducted in Hua County, Henan Province—a region with historically high ESCC incidence—the trial enrolled 34,000 participants aged 45–69 years between 2012 and 2016. This nested study leveraged ESECC’s longitudinal design to investigate whether familial predisposition alters lipid-associated cancer risk, hypothesizing that intrinsic genetic factors might explain heterogeneous associations observed in prior population studies.
Methods
Study Design and Population
From the ESECC cohort, 211 incident cases of malignant esophageal lesions (severe dysplasia, carcinoma in situ, or ESCC) diagnosed via endoscopic screening or follow-up through May 2018 were identified. Each case was matched with up to 10 controls (total 2,101) using incidence density sampling, stratified by age (5-year groups), sex, family history of EC, and screening participation status. Rigorous quality control excluded individuals with incomplete questionnaires or invalid lipid measurements.
Lipid Measurement and Covariates
Fasting serum samples collected at enrollment were analyzed for total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) using a Hitachi 7600 analyzer. Lipid values were converted to mmol/L (TC/LDL/HDL: ×0.0259; TG: ×0.0113). Baseline covariates included BMI, heating fuel type (coal vs. others), kitchen fume exposure, eating speed, and leftover consumption—factors previously linked to ESCC risk in this population.
Statistical Analysis
Conditional logistic regression models evaluated associations between lipid levels and malignant lesions, adjusted for covariates. Two analytical approaches were employed:
- Cutoff Shifting: Continuous lipid measures were dichotomized at increments from the 5th to 95th percentiles of control distributions. Odds ratios (ORs) for high vs. low levels were calculated across cutoff points to assess consistency.
- Quintile Analysis: Lipid values were categorized into quintiles based on control distributions, with the lowest quintile as reference.
Effect modification was tested by introducing interaction terms between lipid indicators and family history of EC, BMI, smoking, and alcohol use. Stratified analyses followed when interaction terms reached statistical significance (P<0.05). Optimal cutoff points maximizing risk heterogeneity between subgroups were selected using Akaike Information Criterion (AIC). Sensitivity analyses excluded prevalent cases diagnosed within 12 months of lipid measurement and participants with cardiovascular disease or diabetes to address reverse causation and medication confounding.
Results
Baseline Characteristics
Cases and controls showed comparable age, sex, and family history distributions. However, cases had significantly lower median BMI (24.09 vs. 24.80 kg/m²; P=0.018). No differences emerged in coal stove use, kitchen fumes, eating speed, or leftover consumption. Median lipid levels were similar between groups: TC (cases: 185 mg/dL vs. controls: 189 mg/dL), TG (113 vs. 117 mg/dL), LDL-C (96 vs. 97 mg/dL), and HDL-C (52 vs. 53 mg/dL) (P>0.05 for all).
Overall Associations
Pooled analysis revealed no consistent associations between lipid parameters and malignancy risk across cutoff points. LDL-C showed marginal inverse association at 100 mg/dL (OR=0.69, 95% CI: 0.48–0.98), but significance dissipated with cutoff adjustments.
Stratification by Family History
Family history of EC emerged as a critical effect modifier:
- High TC (≥200 mg/dL) increased risk among those with familial predisposition (OR=2.22, 95% CI:1.14–4.35) but was protective in non-familial subjects (OR=0.69, 95% CI:0.48–0.98; interaction P=0.002).
- High LDL-C (≥110 mg/dL) similarly elevated risk in familial groups (OR=1.93, 95% CI:1.01–3.65) while reducing risk elsewhere (OR=0.50, 95% CI:0.34–0.76; interaction P<0.001).
TG and HDL-C showed no significant interactions. Optimal cutoffs identified via AIC aligned with clinical thresholds for borderline dyslipidemia (TC:200 mg/dL; LDL-C:110 mg/dL). Sensitivity analyses excluding early-onset cases and comorbidities confirmed robustness of these findings.
Discussion
This study provides novel evidence that familial EC susceptibility dramatically alters serum lipid associations with esophageal malignancy. The bidirectional effects of TC and LDL-C—risk-enhancing in genetically predisposed individuals versus protective in others—explain null findings in prior unstratified analyses. These results underscore the necessity of considering genetic modifiers in lipid-cancer epidemiology.
Mechanistic Insights
Familial aggregation of ESCC likely reflects inherited variations in lipid metabolism pathways. Genome-wide studies implicate loci like LPIN2 (involved in TG synthesis) in familial EC susceptibility. High LDL/TC may accelerate carcinogenesis in genetically vulnerable epithelia through oxidative stress or membrane disruption. Conversely, in low-risk individuals, elevated lipids could exert antitumor effects via cholesterol-mediated apoptosis or enhanced immune surveillance. Nutrition-gene interactions warrant exploration, particularly in regions with dietary lipid variability.
Clinical and Preventive Implications
Family history assessment could refine ESCC risk stratification. For screening programs, lipid profiles may aid personalized risk prediction when integrated with familial data. The convergence of optimal lipid cutoffs with cardiovascular guidelines suggests shared metabolic pathways between EC and CVD in high-risk subgroups, urging interdisciplinary preventive strategies.
Strengths and Limitations
Key strengths include the prospective design, population-based sampling, and pre-diagnostic lipid measurements minimizing reverse causation. Limitations include restricted sample size for subgroup analyses and absence of genetic data to directly test lipid-gene interactions. Generalizability to non-Chinese populations requires validation.
Conclusion
This nested case-control study demonstrates that family history of EC critically modifies serum lipid associations with esophageal malignancy. High TC and LDL-C increase risk in genetically susceptible individuals while exerting protective effects in others, highlighting the importance of stratified approaches in cancer epidemiology. Future research should delineate molecular mechanisms underlying this interaction and evaluate lipid-modulating interventions in high-risk families.
doi.org/10.1097/CM9.0000000000001432
Was this helpful?
0 / 0