Abstract
Background
While it is commonly observed that congenital malformations (CM), birth weight and neuroblastoma (NB) often co-occur clinically, there is a scarcity of studies investigating their relationship.
Aim
To investigate the causal relationship between CM, birth weight and NB using two-sample mendelian randomization (MR) analysis.
Methods
This study utilized data from the Genome-Wide Association Studies (GWAS) database for CM, birth weight, and NB. We identified and analyzed single nucleotide polymorphisms (SNPs) related to CM from various organ systems, ensuring robust instrumental variables through linkage disequilibrium (LD) and F-statistic testing. SNPs were further validated in GWAS Catalog to exclude weak variables. We employed various MR methods including inverse variance weighted (IVW), MR-Egger regression, weighted median, simple mode and weighted mode to assess the causal relationships. Sensitivity analyses were performed using leave-one-out (LOO) and MR-PRESSO.
Results
The analysis demonstrated a protective effect of genital organ CM, which were inversely associated with both overall neuroblastoma (OR = 0.71, 95% CI: 0.60–0.85, P = 0.001) and the 11q-deleted subtype (OR = 0.33, 95% CI: 0.13–0.85, P = 0.02). In contrast, urinary system,gallbladder, bile ducts, and liver were positively associated with neuroblastoma harbouring 11q deletion (OR = 2.37, 95% CI: 1.32–4.27, P = 0.01),(OR = 1.47, 95% CI: 1.17–1.84 P = 0.001) but not with NB overall.However, no significant associations were found between other CM types, birth weight, and NB, and no evidence of heterogeneity, horizontal pleiotropy, or outlier SNPs was identified.
Conclusion
Genital organ malformations confer protection against both overall and 11q-deleted neuroblastoma, whereas urinary system, gallbladder, biliary and liver malformations increase risk exclusively for 11q-deleted disease. However, due to the relaxed p-value threshold used for instrument selection, these findings should be interpreted cautiously and warrant replication with stricter criteria.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12887-025-06295-z.
Keywords: Congenital malformations, Birth weight, Neuroblastoma, Causality, Mendelian randomization
Introduction
Congenital malformations (CM), also known as birth defects, are a group of disorders that result in structural, morphological, functional, behavioral, metabolic, and mental abnormalities caused by external and internal factors during embryonic development. It has been reported [1] that there are about 8 million CM in the world every year, accounting for about 6% of the total number of newborns in the world, and the proportion of children under 5 years of age who die of CM is as high as 8%, which seriously affects the survival and quality of life of children, the happiness and harmony of families, and also creates a huge socio-economic burden [2]. It is now widely recognized that a complex interplay between genetic and environmental influences determines the nature and severity of most CM [3]. In recent years, advances in genomics and genetics have provided new avenues for a deeper understanding of the mechanisms by which CM occur. Through the application of these technologies, we are able to better identify and understand the genetic variants associated with CM, thus providing more precise methods for prevention and treatment [4].
Neuroblastoma (NB), an embryonic tumor arising from sympathetic nervous tissue, is one of the most common cancers in early infancy, accounting for 13% of all deaths from childhood malignancies [5]. The etiology of NB is currently unknown, and the limited research available suggests that it originates from neural crest progenitor cells and is most commonly found in the abdomen, especially the adrenal glands [6]. In recent years, the Children's Oncology Group (COG) has categorized NB into three risk classes, low-risk, intermediate-risk, and high-risk groups [7]. Among them, the long-term survival rate of the highly differentiated group is not optimistic. Improving the survival rate of high-risk NB patients is an urgent clinical problem. It is necessary to further explore new effective targets to diagnose and treat NB.
Previous studies, such as a large French case–control study, have highlighted a significant association between children born with CM and the later development of NB [8]. Genetic analyses have shown that children with CM are 11.6 times more likely to be diagnosed with cancer compared to those without CM, with non-chromosomal CM particularly associated with a higher risk of hepatoblastomas and NB [8, 9]. However, the exact relationship between CM and NB remains debated, with no consensus on which specific types of CM contribute to NB risk. Moreover, while a large meta-analysis has suggested a correlation between high birth weight and embryonic neoplasia, some argue that fetal growth itself, rather than birth weight, may be the primary factor influencing cancer risk. Despite these observations, biological mechanisms linking fetal growth abnormalities with NB risk remain speculative. To address these uncertainties, traditional epidemiological studies have provided valuable insights, yet they are limited by confounding factors and reverse causality. Genome-wide association studies (GWAS) offer a more robust genetic perspective, providing an opportunity to better understand the genetic architecture underlying these relationships. This study applies Mendelian randomization (MR) methods, which use genetic variants as instrumental variables to assess causal relationships while mitigating confounding and reverse causality issues [9]. MR offers a methodological advantage over traditional observational studies by providing a more reliable assessment of causal inference, making it possible to draw more accurate conclusions about the interplay between CM, birth weight, and NB. By integrating GWAS data at the genetic level, this study aims to offer new insights into the causal pathways between these factors, with the potential to inform preventive strategies and improve risk stratification for NB in clinical practice.
Methods
Data sources
CM GWAS summary statistics from FinnGen (https://www.finngen.fi/en) consortium R12 release data [8]. Our analysis incorporates GWAS data encompassing a wide range of body parts and systems. This includes cardiovascular, musculoskeletal, ocular, auditory, facial, and cervical CM, as well as reproductive, nervous, respiratory, urinary, gallbladder, bile duct, and hepatic CM. Additionally, we include other CM related to the digestive system. GWAS summary statistics for birth weight are from data published by The Early Growth Genetics (EGG) Consortium (https://egg-consortium.org) [10], almost all samples were from European populations. Given the significant differences in genetic mechanisms, clinical manifestations, and pathogenesis between congenital metabolic defects and congenital malformations, this study utilized data on congenital malformations from the 17th edition of the International Classification of Diseases (ICD), while excluding data on metabolic defects to conduct a Mendelian randomization analysis. The GWAS datasets selected for this research were noted for their large sample sizes, high data quality, and strong alignment with the research objectives. The FinnGen project provides genetic and health data for over 500,000 Finnish individuals, yielding a robust sample size and genetic homogeneity, which mitigates the confounding effects of population stratification and enhances the precision of genetic association findings. The EGG Consortium focuses on early growth traits, aggregating data from multiple cohorts. These two datasets collectively offer a rich sample size, substantial genetic variability, and detailed phenotypic information, rendering them suitable for investigating the genetic influences on disease and early growth characteristics.
GWAS summary statistics for overall neuroblastoma were retrieved from the IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/); these data correspond to the European-ancestry meta-analysis originally published by Capasso et al. [11]. Summary statistics for the 11q-deletion neuroblastoma subtype were downloaded from the NHGRI-EBI GWAS Catalog (https://www.ebi.ac.uk/gwas/), based on the GWAS reported by Chang et al. [12]. Our analyses used only the publicly released summary-level files;no individual-level genotypes were accessed.The basics of the study and data extraction are presented in (Table 1).
Table 1.
Database data summary information
| Trait | Data sources | Racial origin of the sample | cases | controls | Sample size | |
|---|---|---|---|---|---|---|
| Exposure | Congenital malformations of gallbladder, bile ducts and liver | FinnGen R12 | European | 5918 | 494,430 | 500,348 |
| Other congenital malformations of the digestive system | FinnGen R12 | European | 3646 | 496,702 | 500,348 | |
| Congenital malformations of the urinary system | FinnGen R12 | European | 2617 | 497,731 | 500,348 | |
| Congenital malformations of ovaries, fallopian tubes and broad ligaments | FinnGen R12 | European | 315 | 498,691 | 499,006 | |
| Congenital malformations of the circulatory system | FinnGen R12 | European | 3957 | 496,391 | 500,348 | |
| Congenital malformations of uterus and cervix | FinnGen R12 | European | 823 | 499,525 | 500,348 | |
| Congenital malformations of the respiratory system | FinnGen R12 | European | 2215 | 278,792 | 281,007 | |
| Congenital malformations of the nervous system | FinnGen R12 | European | 585 | 499,763 | 500,348 | |
| Congenital malformations of genital organs | FinnGen R12 | European | 2168 | 498,180 | 500,348 | |
| Congenital malformations of eye, ear, face and neck | FinnGen R12 | European | 776 | 278,792 | 279,568 | |
| Congenital malformations and deformations of the musculoskeletal system | FinnGen R12 | European | 1657 | 498,691 | 500,348 | |
| Birth weight | Early Growth Genetics Consortium | European | NA | NA | 143,677 | |
| Outcome | Neuroblastoma | IEU OpenGWAS | European | 1627 | 3254 | 4881 |
| Neuroblastoma (11q deletion) | GWAS Catalog | European | 113 | 5109 | 5222 |
Study design
This study utilized publicly available GWAS data in a two-sample MR study. Informed consent and ethical approval was obtained from the original publication and the above mentioned databases prior to the commencement of the study. MR analyses are based on three key assumptions: i) The instrumental variables are strongly associated with congenital malformations and birth weight; ii) the instrumental variables are independent of any confounding factors related to neuroblastoma; and iii) the instrumental variables influence the outcome (i.e., the incidence of neuroblastoma) only through the exposure of birth weight or congenital malformations. To address the challenges posed by pleiotropy in MR analysis, this study employed MR-Egger regression, which effectively reduces bias and enhances the accuracy of causal inference (Fig. 1).
Fig. 1.
The study design of two-sample MR analysis
Selection of tool variables
First, single nucleotide polymorphisms (SNPs) closely associated with exposure were identified, and SNPs with P-values less than 5 × 10^−8 were selected. Given the limited number of SNPs of genome-wide significance, a more relaxed threshold (P < 5 × 10^−5) was also used to identify SNPs. Second, in order to obtain independent SNPs, based on the reference genome data of the European population from the Human Genome Project [13], SNPs were collected under a linkage disequilibrium (LD) r2 threshold of r2 < 0.001 and kb > 10,000. Third, the strength of the instrumental variable was tested by calculating the F statistic, which is given by F = R^2 × (N—2)/(1—R^2) [14]. Where N represents the sample size and R^2 represents the proportion of exposed variation explained by the selected SNPs [15, 16]. SNPs with F-statistics greater than 10 were selected to minimize bias in instrumental variables. Fourth, GWAS Catalog was used to assess whether the screened SNPs were confounders. Fifth, SNPs directly and significantly associated with NB (P < 5 × 10^−6) were excluded to fulfill the exclusion assumption.
Statistical analyses
The analysis employed several MR methods, including Inverse Variance Weighting (IVW), MR-Egger, Weighted Median, Simple Modeand Weighted Mode (Figs. 2 and 3). Among these methods, the IVW technique is particularly noted for its frequent application in MR studies; it integrates the results of multiple genetic variants to estimate causal effects. Each genetic variant provides an estimate of this effect, and the estimates are weighted based on their precision (i.e., the magnitude of the standard error) [17]. The higher the precision of the estimate (i.e., the smaller the standard error), the greater the assigned weight. This method operates under the assumption that all genetic variants are valid causal instruments.
Fig. 2.
Heatmap of birth weight and congenital malformations in children with neuroblastom (11q deletion)
Fig. 3.
Heatmap of birth weight and congenital malformations in children with neuroblastom
MR-Egger serves as an extension of the IVW method, aiding in the detection and consideration of potential genetic effect bias (pleiotropic effects), wherein a genetic variant may influence multiple traits, potentially resulting in skewed outcomes [18]. If the results from the MR-Egger method differ from those of the IVW method, this discrepancy may indicate the presence of genetic effect bias. This approach facilitates the estimation of causal effects even when potential bias is present. The odds ratios (OR) and 95% confidence intervals (CI) of these methods were also visualized.
Sensitivity analysis
We performed Cochran's Q test to analyze the heterogeneity and if the P-value obtained was less than 0.05, it indicated the presence of heterogeneity. We assessed horizontal multiplicity by MR-Egger intercept test to check whether the intercept obtained from the estimation deviates from 0. If it does, there may be horizontal multiplicity. In addition, we performed a sensitivity assessment using leave-one-out (LOO) analysis to speculate on the presence or absence of outliers, and used the MR-PRESSO method to detect and assess the impact of outliers on the results. All MR analysis methods available used the TwoSampleMR package (version 0.5.6) in R 4.3.0 software.
Results
Selection of instrumental variables
In accordance with the above selection criteria, the first step was to perform a linkage disequilibrium test to screen for SNPs in which CM, birth weight were linked to NB, SNPs with F-statistics below 10 were then removed. No confounders were seen based on the GWAS Catalog database. Finally, the MR-PRESSO outlier test was conducted with the number of distributions set to 1000. The results indicated P-values greater than 0.05, suggesting that no outliers were detected. Consequently, we did not exclude any SNPs based on the aforementioned selection criteria.Detailed statistics are shown in (Supplementary Table 1,Supplementary Table 2).
Two-sample MR analysis
Association between genital organ CM and NB
We employed the IVW method as the primary approach for MR to evaluate the causal relationships between CM, birth weight, and NB. Notably, genital organ CM was found to be negatively associated with NB harbouring (11q deletion) (OR = 0.33, 95% CI: 0.13–0.85, P = 0.02), as illustrated in (Fig. 4, Supplementary Figs. 3G) and with neuroblastoma overall (OR = 0.71, 95% CI: 0.60–0.85, P = 0.001) (Fig. 5 and Supplementary Figs. 4G).The scatter plot emphasizes the strong correlation between the genetic instruments and their effects on the risks of genital organ CM and NB, supporting the validity of the instruments used in the MR analysis and ensuring the robustness of the causal inference. To assess the stability of the results, we conducted a Cochrane Q test on the MR outcomes derived from the IVW method. This analysis revealed no significant heterogeneity in the causal relationship between genital organ CM and NB (P > 0.05). Furthermore, MR-Egger regression was employed to evaluate potential horizontal pleiotropy between the instrumental variables and the outcomes. The Egger intercept test indicated no evidence of horizontal pleiotropy (P > 0.05) (Tables 2 and 3). Additionally, LOO sensitivity analysis revealed no bias attributable to any single SNP (Supplementary Fig. 1G, Supplementary Fig. 2G), further confirming the reliability of the observed causal relationship in our analysis.
Fig. 4.
MR estimates of the relationship of genetically predicted birth weight and congenital malformations on neuroblastoma(11q deletion)
Fig. 5.
MR estimates of the relationship of genetically predicted birth weight and congenital malformations on neuroblastoma
Table 2.
MR analysis of birth weight, congenital malformations and neuroblastoma(11q deletion): heterogeneity and pleiotropy test results (P < 5 × 10–5)
| Exposure | Method | No.of SNP | Heterogeneity Q | Heterogeneity P | Egger Intercept | Egger Intercept P |
|---|---|---|---|---|---|---|
| Birth weight | Inverse variance weighted | 8 | 10.038 | 0.186 | ||
| MR Egger | 0.004 | 0.985 | ||||
| Congenital malformations and deformations of musculoskeletal system | Inverse variance weighted | 7 | 9.589 | 0.143 | ||
| MR Egger | 0.086 | 0.806 | ||||
| Congenital malformations of eye, ear, face and neck | Inverse variance weighted | 7 | 1.762 | 0.940 | ||
| MR Egger | −0.198 | 0.546 | ||||
| Congenital malformations of gallbladder, bile ducts and liver | Inverse variance weighted | 3 | 0.818 | 0.664 | ||
| MR Egger | −0.154 | 0.584 | ||||
| Congenital malformations of genital organs | Inverse variance weighted | 6 | 3.951 | 0.556 | ||
| MR Egger | 0.476 | 0.196 | ||||
| Congenital malformations of ovaries, fallopian tubes and broad ligaments | Inverse variance weighted | 10 | 3.614 | 0.935 | ||
| MR Egger | 0.049 | 0.795 | ||||
| Congenital malformations of the circulatory system | Inverse variance weighted | 6 | 4.75 | 0.447 | ||
| MR Egger | 0.099 | 0.751 | ||||
| Congenital malformations of the nervous system | Inverse variance weighted | 5 | 0.732 | 0.947 | ||
| MR Egger | −0.114 | 0.576 | ||||
| Congenital malformations of urinary system | Inverse variance weighted | 8 | 4.678 | 0.699 | ||
| MR Egger | 0.235 | 0.426 | ||||
| Congenital malformations of uterus and cervix | Inverse variance weighted | 3 | 1.335 | 0.513 | ||
| MR Egger | 0.406 | 0.486 | ||||
| Other congenital malformations of the digestive system | Inverse variance weighted | 7 | 4.285 | 0.638 | ||
| MR Egger | 0.132 | 0.623 |
Table 3.
MR analysis of birth weight, congenital malformations and neuroblastoma: heterogeneity and pleiotropy test results (P < 5 × 10–5)
| Exposure | Method | No.of SNP | Heterogeneity Q | Heterogeneity P | Egger Intercept | Egger Intercept P |
|---|---|---|---|---|---|---|
| Birth weigh | Inverse variance weighted | 8 | 3.312 | 0.855 | ||
| MR Egger | −0.017 | 0.734 | ||||
| Congenital malformations and deformations of musculoskeletal system | Inverse variance weighted | 6 | 8.031 | 0.155 | ||
| MR Egger | 0.058 | 0.632 | ||||
| Congenital malformations of eye, ear, face and neck | Inverse variance weighted | 7 | 4.25 | 0.643 | ||
| MR Egger | 0.045 | 0.657 | ||||
| Congenital malformations of gallbladder, bile ducts and liver | Inverse variance weighted | 2 | 0.224 | 0.636 | ||
| Congenital malformations of genital organs | Inverse variance weighted | 10 | 6.529 | 0.686 | ||
| MR Egger | −0.054 | 0.279 | ||||
| Congenital malformations of ovaries, fallopian tubes and broad ligaments | Inverse variance weighted | 9 | 13.44 | 0.098 | ||
| MR Egger | −0.002 | 0.985 | ||||
| Congenital malformations of the circulatory system | Inverse variance weighted | 5 | 1.695 | 0.792 | ||
| MR Egger | −0.128 | 0.319 | ||||
| Congenital malformations of the nervous system | Inverse variance weighted | 5 | 3.023 | 0.554 | ||
| MR Egger | 0.082 | 0.269 | ||||
| Congenital malformations of urinary system | Inverse variance weighted | 8 | 3.685 | 0.815 | ||
| MR Egger | 0.056 | 0.567 | ||||
| Congenital malformations of uterus and cervix | Inverse variance weighted | 3 | 3.529 | 0.171 | ||
| MR Egger | −0.225 | 0.332 | ||||
| Other congenital malformations of the digestive system | Inverse variance weighted | 6 | 2.777 | 0.734 | ||
| MR Egger | 0.041 | 0.641 |
Analysis of birth weight, remaining CM, and NB
Using the IVW method as the primary Mendelian randomization analysis, we evaluated the causal relationships between CM, birth weight, and neuroblastoma. Genetically predicted malformations of the urinary system,gallbladder, bile ducts, and liver were positively associated with neuroblastoma harbouring 11q deletion (OR = 2.37, 95% CI: 1.32–4.27, P = 0.01),(OR = 1.47, 95% CI: 1.17–1.84 P = 0.001) (Fig. 4, Supplementary Figs. 3 C,H). Although the causal estimates for genitourinary, gallbladder, biliary-duct and hepatic malformations with 11q-deleted neuroblastoma exhibited no significant heterogeneity (P > 0.05) or horizontal pleiotropy (P > 0.05) (Table 2), these malformations showed no overall causal association with neuroblastoma. Consequently, these findings should be interpreted with caution. Furthermore, no significant associations were detected between any other congenital malformation subtypes or birth weight and either neuroblastoma overall or neuroblastoma with 11q deletion. The results indicated no significant correlation among the remaining CM, birth weight, and NB, as illustrated in the forest plot and scatter plot (Figs. 4 and 5, Supplementary Figs. 3 A,B,C,D,E,H,I,J,K; Supplementary Figs. 4 A,B,D,E,I,J,K).
The heterogeneity of the MR results was assessed using the IVW method, which revealed no significant heterogeneity (Tables 2 and 3). Horizontal pleiotropy was evaluated through MR-Egger regression, and the Egger intercept test confirmed the absence of horizontal pleiotropy (P > 0.05) (Tables 2 and 3). Although Cochran's Q-test and MR-Egger analyses did not detect heterogeneity or pleiotropy among single nucleotide polymorphisms (SNPs) (P > 0.05), the LOO analysis for birth weight, remaining CM suggested the potential for an influential SNP-driven causal relationship with NB (Supplementary Figs. 1A,B,D,E,F, I,J,K; Supplementary Figs. 2A,B,D, E,F, I,J).
Furthermore, no eligible SNPs were identified for congenital malformations of the respiratory system. In addition, the instrumental variables for gallbladder, biliary-duct and liver malformations exhibited insufficient power to support sensitivity analyses (Table 3). These limitations underscore the need for cautious interpretation of the present findings.
Following Bonferroni correction (α = 0.0045), the inverse association between genital organ CM and overall NB (P = 0.001) and the positive association between gallbladder, bile ducts and liver malformations and 11q-deleted neuroblastoma remained significant. In contrast, the associations of urinary system malformations (P = 0.01) and genital organ malformations (P = 0.02) with 11q-deleted neuroblastoma did not reach the corrected threshold.
Discussion
NB is one of the most common extracranial solid tumors in children [19] and the most common type of cancer among infants up to 12 months of age [20], occurring almost exclusively in children and accounting for approximately 15% of all childhood cancer deaths. Its incidence is 107 cases per 1 million children diagnosed with NB [21]. In addition the onset of NB is age-related with a median age at diagnosis of 17.3 months, with 40% of patients diagnosed before the age of 1 year [22]. The exact risk factors for NB have not been defined so far. Some reports suggest that maternal smoking [23], opioid use during pregnancy or lactation [24], gestational diabetes mellitus [25], and insufficient folic acid intake [26] may be risk factors influencing the onset and progression of this disease. Although many maternal and fetal exposures, as well as genetic disorders, may sometimes accompany NB, the exact pathogenic mechanisms remain unclear.
Drawing from our comprehension, our study marks the first application of MR methods to comprehensively investigate the causal impact of CM, birth weight in relation to NB. Compared with previous observational studies exploring the relationship between CM, birth weight and NB, the use of an MR design reduces the potential for confounding variables and bias. For example, some studies have indicated that birth weight is significantly associated with different types of cancer in infants and children [27, 28], but there is insufficient evidence for NB. A meta-analysis by Olgun HN [29] indicated that either high or low birth weight may be an important risk factor for NB, but it is important to note that interview-based studies may be relatively biased. Although genetic susceptibility to CM and birth weight may subtly influence neurological alterations, the results of this study indicate that no significant association was found between birth weight and other congenital malformations or NB. This finding is perplexing, given that previous studies have indicated that low birth weight (< 2500 g) may reduce the risk of neuroblastoma while high birth weight is associated with specific congenital malformations [30]. This result is inconsistent with our findings. After a comprehensive analysis, we consider that the reasons for this discrepancy may be related to the database we selected. Some studies have indicated that maternal and fetal genetic effects may independently influence birth weight [31]. In GWAS databases, maternal and fetal genetic effects related to birth weight overlap, and our study did not distinguish between these effects, which could potentially lead to errors in the results. Of course, it is also possible that intrauterine factors or fetal factors themselves influenced the results of this study, as previous research has suggested that intrauterine factors are also crucial determinants for the child [9]. Finally, the choice of the sample population may also contribute to the findings of this study. Our research predominantly focuses on the European population, and selection bias could exist. In addition, we performed sensitivity analyses to ensure the validity of our hypothesis.
Children with CM are more likely to develop cancer in childhood or later in life. Recent studies have for the first time explored the genomes of children carrying CM, with or without concomitant cancer, and found differentially expressed protein-coding/non-coding RNAs between the two phenotypes, which may indicate that genetic factors underlie the molecular biology of cancer in children with CM [32]. Liu's team [30] analyzed the genomes of 1,653 chromosome-abnormality-free individuals by blood DNA whole genome sequencing (WGS), identified genetic variants associated with malignancies coexisting with CM. These studies further revealed the underlying molecular mechanisms of cancer in children with CM.
While there are indications that children with CM are predisposed to cancer, the specific association of NB is unclear. Previous studies have suggested that severe congenital anomalies may be associated with the development of NB, particularly with genitourinary and cardiac malformations, and that this association is more pronounced in infants under 1 year of age [33–35]. Our MR analysis corroborates and refines this observation: genetically predicted urogenital malformations were positively associated with neuroblastoma harbouring 11q deletion, indicating that at least a subset of CMs confers susceptibility to distinct NB molecular subtypes. However, other studies failed to confirm this finding [36–38]. Although reports on the association between congenital heart disease and NB exist [28], the conclusions remain inconclusive. A potential problem with these studies is that children with cancer may also be found to carry CM during diagnostic evaluation, making it difficult to establish a causal relationship. To address this limitation, future research should focus on increasing the number of patients and samples studied. A larger dataset would allow for more robust statistical analyses and potentially reveal subtle associations that are not apparent in smaller cohorts. Strengthening the sample size could provide more definitive insights into how congenital factors might contribute to neuroblastoma risk, thereby enhancing our understanding and guiding clinical practices more effectively.
This study concluded that congenital malformations of the genital organs serve as a protective factor against NB, while the remaining congenital malformations demonstrate no causal relationship with NB. A potential mechanism may involve speculating on the experimental results concerning estrogen receptor alpha (ERα) and its influence on neuroblastoma progression, which suggests the possibility of an indirect connection. For instance, ERα overexpression in neuroblastoma cells induces differentiation and disrupts tumorigenic processes, including reduced proliferation and anchorage-independent growth. This implies that elevated levels of ERα may promote a less aggressive and more differentiated tumor phenotype, potentially correlating with protective factors in other tissues or malformations, particularly those of the genitourinary system [39].
Genitourinary malformations might also influence the expression or regulation of certain signaling pathways, such as those involving growth factors (e.g., NGF) or hormonal regulation (e.g., estrogen signaling) [40]. The crosstalk between these pathways could potentially interfere with the tumorigenic processes in neuroblastoma, making the tumor less likely to develop or progress in individuals with certain genitourinary anomalies. We speculated that ERα signalling might underlie the apparently protective effect of genital-organ malformations on NB risk. While in-vitro data indicate that ERα over-expression can induce differentiation and suppress proliferation of NB cells, these observations remain correlative. The precise molecular pathway linking genital-organ malformations to ERα activity in NB has not been experimentally tested. Therefore, we explicitly frame this as a hypothesis that requires functional validation through targeted knock-down/knock-in studies and organoid models before any causal mechanism can be asserted.
To draw a concrete conclusion about this relationship, more research would be needed to explore how genitourinary malformations impact the molecular and cellular biology of neuroblastoma, particularly through hormonal or signaling pathway modulation.
However, exploring the causal relationship between CM and NB can be complex because both conditions are relatively rare, and their etiologies can be multifactorial, involving a combination of genetic, environmental, and developmental factors. Several potential mechanisms can be proposed to explain the association between CM and NB. Some genetic syndromes that cause birth defects also increase the risk of developing NB: ①Epigenetic mechanisms, play a crucial role in the regulation of gene expression during development. Abnormal epigenetic regulation can lead to both CM and NB. For example, a study identified a common DNA methylation in individuals with recurrent constellations of embryonic malformations (RCEM), which includes multiple malformations affecting three or more body structures. This suggests that epigenetic dysregulation may be a common pathway underlying the development of CM and NB. Epigenetic changes can be induced by environmental factors such as maternal exposure to certain substances or medications, which may explain the shared risk factors for CM and NB. ②For example, mutations in the PHOX2B gene [41] are associated with both congenital central hypoventilation syndrome and Hirschsprung disease.These mutations can disrupt the normal development of neural crest cells, leading to both structural CM and an increased risk of NB. ③Both CM and NB may have origins in early embryonic development. Neural crest cells, which give rise to NB, are also involved in the development of various tissues and organs, including the heart, gut, and nervous system. Abnormal development of these cells can lead to structural CM and cancer. For example, congenital heart defects (CHDs) are one of the most common types of CM and can be associated with disruptions in the development of neural crest cells [42]. Similarly, Hirschsprung disease, which is associated with mutations in the PHOX2B gene, also involves neural crest cell migration and differentiation [43]. This shared embryonic origin suggests that disruptions in the early development of neural crest cells may increase the risk of both CM and NB. ④Maternal Health and Environmental Factors Maternal health conditions and environmental exposures during pregnancy can increase the risk of both CM and NB. For example, maternal alcohol consumption during pregnancy has been linked to a range of congenital anomalies, including CHDs, cleft lip and/or palate, and limb deficiencies [44]. Maternal exposure to certain medications, such as anticonvulsants, or environmental pollutants, such as perfluoroalkyl substances (PFAS), has also been associated with an increased risk of CHDs [43]. These exposures can disrupt normal embryonic development and increase the risk of both CM and NB. Different types of CM may have distinct underlying causes, and not all of them may be related to birth weight. However, the potential pathways involving epigenetic regulation, shared embryonic origins, genetic syndromes, and maternal health and environmental factors provide a framework for understanding the complex relationship between CM and NB. Future research should aim to further explore these pathways and their interactions to better understand the causal relationship between CM and NB.
Nonetheless, the bulk of prior research has been observational, leaving the causal link uncertain. In contrast to earlier conjectures, except from CM of the reproductive organs, our MR analysis fails to substantiate a causal association between CM, birth weight and NB. One of the advantages of MR analysis is that genetic variations used as instrumental variables are randomly assigned during meiosis, which is a process analogous to a randomized controlled trial in genetics [45]. This random assignment at conception means that the genetic variants which are associated with the risk factor of interest are distributed independently of confounding factors that typically bias observational studies. Since genetic variants are fixed at conception, they can provide a measure of the risk factor that is lifelong. This allows MR studies to assess the cumulative effect of a risk factor over a lifespan, rather than at a single point in time, which might not reflect long-term exposure. In observational studies, unmeasured confounding can bias the estimates of the effect of a risk factor on an outcome [46]. MR relies on genetic variants that are not typically associated with confounding factors, thereby offering an opportunity to estimate the causal effect of the risk factor on an outcome more reliably. Furthermore, the substantial sample size constitutes another advantage of this study, mitigating sampling discrepancies. Future investigations are warranted to illuminate the underlying factors that could have contributed to bias or confounding in preceding observational studies.
However, it is also important to acknowledge that MR is not without its limitations and assumptions. One key assumption is the absence of pleiotropy, where a genetic variant affects the outcome through multiple pathways and not just through the risk factor of interest. Another limitation can be the strength of the association between genetic variants and the risk factor, as weak instruments can lead to biased MR estimates. Additionally, MR studies require large sample sizes to achieve sufficient statistical power, especially when the genetic effect on the risk factor is small. Fourthly, while Cochran's Q test and MR-Egger analysis demonstrated robustness among each SNP, the presence of outliers and skewed distribution in funnel plots suggests the possibility of residual heterogeneity and pleiotropy in the IVs [47].
Additionally, the limitations of our study also include: ① Our study is based on Mendelian randomization, and as mentioned in the discussion section, there is some data overlap and genetic effects, which may explain the lack of significant findings in relation to body weight; ② This study is based on a European population, which presents certain limitations. Whether the conclusions are applicable to other racial populations remains to be verified; ③ The issue of genetic pleiotropy cannot be ruled out. Genetic pleiotropy is a potential cause of the observed result bias; ④ A key limitation of this study is the use of a lenient p-value threshold (5 × 10⁻5) for SNP selection, which may increase the risk of correlated horizontal pleiotropy and false-positive causal inference. Although LD-clumping and sensitivity analyses (MR-Egger and MR-PRESSO) were applied, these methods may be underpowered when the number of instrumental variables is limited. Furthermore, advanced approaches such as PCMR and CAUSE could not be reliably implemented due to insufficient SNP counts (< 10). Thus, our findings should be interpreted as exploratory, and validation using larger GWAS datasets or alternative study designs is warranted.
In summary, this study is the first to comprehensively assess the causal relationship between CM and birth weight with NB from a genetic perspective using MR by integrating GWAS data. In contrast to previous observational studies, MR analysis could more effectively contribute to prevention of this disease, irrespective of whether there is an association between CM, birth weight and the risk of developing NB, diagnosis and treatment. The results of the study will not only provide a scientific basis for public health policymakers but will also offer guidance to clinicians in identifying high-risk individuals and providing early screening and intervention. Moreover the study will help improve the understanding of the mechanisms of CM. By exploring the association between CM, birth weight and NB, it will reveal the underlying biological and physiological processes, thus promoting further research in related fields.
Lastly, some experiments being carried out may provide more favorable evidence for further exploring the association between the two. For instance, validating these findings in diverse populations would be crucial to ensure the generalizability of the results. This could involve conducting similar MR analyses in different ethnic and geographic groups to confirm the consistency of the observed associations. Additionally, exploring functional genomics could uncover the underlying mechanisms. This might include investigating the role of specific genetic variants in the development of CM and NB, as well as their interaction with environmental factors. Such research could provide insights into the molecular pathways involved and potentially identify new therapeutic targets. Furthermore, integrating multi-omics data, such as transcriptomics and metabolomics, could offer a more comprehensive understanding of the biological processes at play. These concrete next steps will not only validate the current findings but also pave the way for more targeted and effective interventions in the future.
In conclusion, the significance of this study is to provide us with important information about the relationship between CM, birth weight and the risk of NB, to provide a basis for prevention, diagnosis, and treatment, and to promote further development of related research areas.
Supplementary Information
Acknowledgements
The authors want to acknowledge all of the participants and investigators for contributing and sharing summary-level data on GWAS. The authors acknowledge the participants and investigators of the FinnGen study.
Abbreviations
- CM
Congenital malformations
- NB
Neuroblastoma
- MR
Mendelian randomization
- TSMR
Two Sample MR
- GWAS
Genome-Wide Association Analysis database
- SNPs
Single nucleotide polymorphisms
- LD
Linkage disequilibrium
- IVW
Inverse variance weighted
- LOO
Leave-one-out
- COG
Children's oncology group
- CI
Confidence intervals
- ICD
International Classification of Diseases
Authors’ contributions
SXL and LZ proposed the conception, designed the study and supervised the whole analysis. AA acquired the data and performed statistical analysis. YR and FA drafted the article. MA contributed to the data acquisition and checked the statistical analysis.
Funding
This work was supported by grants from Key R&D Program of Xinjiang (2023B03018).
Data availability
The original contributions presented in the study are included in the article/Supplementary material.further inquiries can be directed to the corresponding author. Publicly available datasets were analyzed in this study. The data can be found here: https://www.finngen.fi/en; https://egg-consortium.org; https://gwas.mrcieu.ac.uk/; https://www.ebi.ac.uk/gwas/.
Declarations
Ethics approval and consent to participate
Approval of the research protocol by an Institutional Reviewer Board: N/A.
Informed consent: N/A.
Registry and the Registration No. of the study/trial: N/A.
Animal studies: N/A.
Consent for publication
Not Applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Ailikamu Aierken, Falide Atabieke, Yierzhati Aizezi contributed equally to this work.
References
- 1.Perin J, Mulick A, Yeung D, et al. Global, regional, and national causes of under-5 mortality in 2000–19: an updated systematic analysis with implications for the sustainable development goals. Lancet Child Adolesc Health. 2022;6(2):106–15. 10.1016/S2352-4642(21)00311-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Apfeld JC, Kastenberg ZJ, Gibbons AT, Phibbs CS, Lee HC, Sylvester KG. The disproportionate cost of operation and congenital anomalies in infancy. Surgery. 2019;165(6):1234–42. 10.1016/j.surg.2018.12.022. [DOI] [PubMed] [Google Scholar]
- 3.Beames TG, Lipinski RJ. Gene-environment interactions: aligning birth defects research with complex etiology. Development. 2020;147(21):dev191064. 10.1242/dev.191064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Khokha MK, Liu KJ, Wallingford JB. Challenges and opportunities at the interface of birth defects, human genetics and developmental biology. Development. 2020;147(21):dev197871. 10.1242/dev.197871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Testori A, Lasorsa VA, Cimmino F, et al. Exploring shared susceptibility between two neural crest cells originating conditions: neuroblastoma and congenital heart disease. Genes. 2019;10(9):663. 10.3390/genes10090663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Maris JM, Hogarty MD, Bagatell R, Cohn SL. Neuroblastoma. Lancet. 2007;369(9579):2106–20. 10.1016/S0140-6736(07)60983-0. [DOI] [PubMed] [Google Scholar]
- 7.Du B, Zhang F, Zhou Q, et al. Joint analysis of the metabolomics and transcriptomics uncovers the dysregulated network and develops the diagnostic model of high-risk neuroblastoma. Sci Rep. 2023;13(1):16991. 10.1038/s41598-023-43988-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kurki MI, Karjalainen J, Palta P, et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. 2023;613(7944):508–18. 10.1038/s41586-022-05473-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ardissino M, Morley AP, Slob EAW, et al. Birth weight influences cardiac structure, function, and disease risk: evidence of a causal association. Eur Heart J. 2024;45(6):443–54. 10.1093/eurheartj/ehad631. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Horikoshi M, Beaumont RN, Day FR, et al. Genome-wide associations for birth weight and correlations with adult disease. Nature. 2016;538(7624):248–52. 10.1038/nature19806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Capasso M, Diskin SJ, Totaro F, et al. Replication of GWAS-identified neuroblastoma risk loci strengthens the role of BARD1 and affirms the cumulative effect of genetic variations on disease susceptibility. Carcinogenesis. 2013;34(3):605–11. 10.1093/carcin/bgs380. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chang X, Zhao Y, Hou C, et al. Common variants in MMP20 at 11q22.2 predispose to 11q deletion and neuroblastoma risk. Nat Commun. 2017;8(1):569. 10.1038/s41467-017-00408-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.1000 Genomes Project Consortium, Abecasis GR, Altshuler D, et al. A map of human genome variation from population-scale sequencing. Nature. 2010;467(7319):1061–73. 10.1038/nature09534. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Burgess S, Thompson SG, CRP CHD Genetics Collaboration. Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol. 2011;40(3):755–64. 10.1093/ije/dyr036. [DOI] [PubMed] [Google Scholar]
- 15.Shim H, Chasman DI, Smith JD, et al. A multivariate genome-wide association analysis of 10 LDL subfractions, and their response to statin treatment, in 1868 Caucasians. PLoS One. 2015;10(4):e0120758. 10.1371/journal.pone.0120758. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chen D, Zhang Y, Yidilisi A, Xu Y, Dong Q, Jiang J. Causal associations between circulating adipokines and cardiovascular disease: a mendelian randomization study. J Clin Endocrinol Metab. 2022;107(6):e2572–80. 10.1210/clinem/dgac048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bowden J, Holmes MV. Meta-analysis and Mendelian randomization: a review. Res Synth Methods. 2019;10(4):486–96. 10.1002/jrsm.1346. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017;32(5):377–89. 10.1007/s10654-017-0255-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Takita J. Molecular basis and clinical features of neuroblastoma. JMA J. 2021;4(4):321–31. 10.31662/jmaj.2021-0077. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gurney JG, Ross JA, Wall DA, Bleyer WA, Severson RK, Robison LL. Infant cancer in the U.S.: histology-specific incidence and trends, 1973 to 1992. J Pediatr Hematol Oncol. 1997;19(5):428–32. 10.1097/00043426-199709000-00004. [DOI] [PubMed] [Google Scholar]
- 21.Matthay KK, Maris JM, Schleiermacher G, et al. Neuroblastoma. Nat Rev Dis Primers. 2016;2:16078. 10.1038/nrdp.2016.78. [DOI] [PubMed] [Google Scholar]
- 22.Castleberry RP. Neuroblastoma. Eur J Cancer. 1997;33(9):1430–8. 10.1016/s0959-8049(97)00308-0. [DOI] [PubMed] [Google Scholar]
- 23.Kramer S, Ward E, Meadows AT, Malone KE. Medical and drug risk factors associated with neuroblastoma: a case-control study. J Natl Cancer Inst. 1987;78(5):797–804. [PubMed] [Google Scholar]
- 24.Cook MN, Olshan AF, Guess HA, et al. Maternal medication use and neuroblastoma in offspring. Am J Epidemiol. 2004;159(8):721–31. 10.1093/aje/kwh108. [DOI] [PubMed] [Google Scholar]
- 25.Chow EJ, Friedman DL, Mueller BA. Maternal and perinatal characteristics in relation to neuroblastoma. Cancer. 2007;109(5):983–92. 10.1002/cncr.22486. [DOI] [PubMed] [Google Scholar]
- 26.French AE, Grant R, Weitzman S, et al. Folic acid food fortification is associated with a decline in neuroblastoma. Clin Pharmacol Ther. 2003;74(3):288–94. 10.1016/S0009-9236(03)00200-5. [DOI] [PubMed] [Google Scholar]
- 27.Hjalgrim LL, Westergaard T, Rostgaard K, et al. Birth weight as a risk factor for childhood leukemia: a meta-analysis of 18 epidemiologic studies. Am J Epidemiol. 2003;158(8):724–35. 10.1093/aje/kwg210. [DOI] [PubMed] [Google Scholar]
- 28.Dahlhaus A, Prengel P, Spector L, Pieper D. Birth weight and subsequent risk of childhood primary brain tumors: an updated meta-analysis. Pediatr Blood Cancer. 2017;64(5). 10.1002/pbc.26299. [DOI] [PubMed]
- 29.Olgun HN, Arayici ME, Cecen RE, Basbinar Y, Ellidokuz H. The association between birth weight and the risk of neuroblastoma: a meta-analysis of observational studies involving 4,361,141 participants. Am J Cancer Res. 2023;13(9):3854–63. [PMC free article] [PubMed] [Google Scholar]
- 30.Rashti R, Ghasemi F, Poorolajal J. Association between birth weight and risk of nonneurological childhood cancers: a systematic review and meta-analysis. Eur J Cancer Prev. 2025;34(1):40–52. 10.1097/CEJ.0000000000000894. [DOI] [PubMed] [Google Scholar]
- 31.Warrington NM, Beaumont RN, Horikoshi M, et al. Maternal and fetal genetic effects on birth weight and their relevance to cardio-metabolic risk factors. Nat Genet. 2019;51(5):804–14. 10.1038/s41588-019-0403-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Liu Y, Qu HQ, Chang X, et al. Genomic variants exclusively identified in children with birth defects and concurrent malignant tumors predispose to cancer development. Mol Cancer. 2023;22(1):126. 10.1186/s12943-023-01828-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Munzer C, Menegaux F, Lacour B, et al. Birth-related characteristics, congenital malformation, maternal reproductive history and neuroblastoma: the ESCALE study (SFCE). Int J Cancer. 2008;122(10):2315–21. 10.1002/ijc.23301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Rios P, Bailey HD, Orsi L, et al. Risk of neuroblastoma, birth-related characteristics, congenital malformations and perinatal exposures: a pooled analysis of the ESCALE and ESTELLE French studies (SFCE). Int J Cancer. 2016;139(9):1936–48. 10.1002/ijc.30239. [DOI] [PubMed] [Google Scholar]
- 35.Menegaux F, Olshan AF, Reitnauer PJ, Blatt J, Cohn SL. Positive association between congenital anomalies and risk of neuroblastoma. Pediatr Blood Cancer. 2005;45(5):649–55. 10.1002/pbc.20263. [DOI] [PubMed] [Google Scholar]
- 36.Buck GM, Michalek AM, Chen CJ, Nasca PC, Baptiste MS. Perinatal factors and risk of neuroblastoma. Paediatr Perinat Epidemiol. 2001;15(1):47–53. 10.1046/j.1365-3016.2001.00307.x. [DOI] [PubMed] [Google Scholar]
- 37.Bluhm E, McNeil DE, Cnattingius S, Gridley G, El Ghormli L, Fraumeni JF Jr. Prenatal and perinatal risk factors for neuroblastoma. Int J Cancer. 2008;123(12):2885–90. 10.1002/ijc.23847. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Neglia JP, Smithson WA, Gunderson P, King FL, Singher LJ, Robison LL. Prenatal and perinatal risk factors for neuroblastoma. A case-control study. Cancer. 1988;61(11):2202–6. 10.1002/1097-0142(19880601)61:11%3c2202::aid-cncr2820611113%3e3.0.co;2-7. [DOI] [PubMed] [Google Scholar]
- 39.Dzieran J, Rodriguez Garcia A, Westermark UK, et al. MYCN-amplified neuroblastoma maintains an aggressive and undifferentiated phenotype by deregulation of estrogen and NGF signaling. Proc Natl Acad Sci USA. 2018;115(6):E1229–38. 10.1073/pnas.1710901115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Zheng Z, Armfield BA, Cohn MJ. Timing of androgen receptor disruption and estrogen exposure underlies a spectrum of congenital penile anomalies. Proc Natl Acad Sci USA. 2015;112(52):E7194–203. 10.1073/pnas.1515981112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Zhao J, Zhu Y, Xie X, et al. Pleiotropic effect of common PHOX2B variants in Hirschsprung disease and neuroblastoma. Aging (Albany NY). 2019;11(4):1252–61. 10.18632/aging.101834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Li Y, Du J, Deng S, et al. The molecular mechanisms of cardiac development and related diseases. Signal Transduct Target Ther. 2024;9(1):368. 10.1038/s41392-024-02069-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Ou Y, Zeng X, Lin S, et al. Gestational exposure to perfluoroalkyl substances and congenital heart defects: a nested case-control pilot study. Environ Int. 2021;154:106567. 10.1016/j.envint.2021.106567. [DOI] [PubMed] [Google Scholar]
- 44.Kurita H, Motoki N, Inaba Y, et al. Maternal alcohol consumption and risk of offspring with congenital malformation: the Japan environment and children’s study. Pediatr Res. 2021;90(2):479–86. 10.1038/s41390-020-01274-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Aierken A, Atabieke F, Aierken M, et al. No bidirectional relationship between inflammatory bowel disease and diverticular disease: a genetic correlation and Mendelian randomization study. Front Genet. 2024;15:1334473. 10.3389/fgene.2024.1334473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Atabieke F, Li XJ, Aierken A, et al. Association between frailty and hepatic fibrosis in NAFLD among middle-aged and older adults: results from NHANES 2017–2020. Front Public Health. 2024;12:1330221. 10.3389/fpubh.2024.1330221. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Richmond RC, Davey SG. Mendelian randomization: concepts and scope. Cold Spring Harb Perspect Med. 2022;12(1):a040501. 10.1101/cshperspect.a040501. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The original contributions presented in the study are included in the article/Supplementary material.further inquiries can be directed to the corresponding author. Publicly available datasets were analyzed in this study. The data can be found here: https://www.finngen.fi/en; https://egg-consortium.org; https://gwas.mrcieu.ac.uk/; https://www.ebi.ac.uk/gwas/.