Anthropometric indicators may explain the high incidence of follicular lymphoma in Europeans: Results from a bidirectional two-sample two-step Mendelian randomisation

doi:10.21203/rs.3.rs-3440102/v1

Download PDF

Article

Anthropometric indicators may explain the high incidence of follicular lymphoma in Europeans: Results from a bidirectional two-sample two-step Mendelian randomisation

https://doi.org/10.21203/rs.3.rs-3440102/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 01 Jun, 2024

Read the published version in Gene →

Version 1

posted

You are reading this latest preprint version

Background

Non-Hodgkin's lymphoma incidence rates vary between European and Asian populations. The reasons remain unclear. This two-sample two-step Mendelian randomisation (MR) study aimed to investigate the causal relationship between anthropometric indicators (AIs) and diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL) and the possible mediating role of basal metabolic rate (BMR) in Europe. (2)

Methods

We used the following AIs as exposures: body mass index (BMI), whole-body fat mass (WBFM), whole-body fat-free mass (WBFFM), waist circumference(WC), hip circumference(HC), standing height (SH), and weight(Wt). DLBCL and FL represented the outcomes, and BMR was a mediator. A two-sample MR analysis was performed to examine the association between AIs and DLBCL and FL onset. We performed reverse-MR analysis to determine whether DLBCL and FL interfered with the AIs. A two-step MR analysis was performed to determine whether BMR mediated the causality. (3) Findings: WBFFM and SH had causal relationships with FL. A causal association between AIs and DLBCL was not observed. Reverse-MR analysis indicated the causal relationships were not bidirectional. Two-step MR suggested BMR may mediate the causal effect of WBFFM and SH on FL. (4)

Conclusions

We observed a causal relationship between WBFFM and SH and the onset of FL in Europeans, Which may explain the high incidence of follicular lymphoma in Europeans.

Health sciences/Oncology/Cancer/Cancer epidemiology

Health sciences/Oncology/Cancer/Haematological cancer/Lymphoma/Non hodgkin lymphoma

Mendelian randomisation

Follicular lymphoma

Anthropometric indicators

Incidence rate

Basal metabolic rate

Non-Hodgkin's lymphoma (NHL), the most prevalent blood cancer worldwide¹, is a diverse group of lymphoproliferative malignancies with varying histological subtypes that is responsible for approximately 3% of all diagnosed cancer cases and fatalities². Although there are over 40 significant subtypes of NHL, most (85.0–90.0%) are of B-cell origin³. Its most prevalent forms include indolent follicular lymphoma (FL) and diffuse large B-cell lymphoma (DLBCL)². DLBCL is the most common type of NHL worldwide, followed by FL⁴. In European countries, DLBCL accounts for 30–40% of all NHL cases^2,5. In Asian countries, such as China, Japan, and South Korea, DLBCL accounts for 40.8%⁶, 36.0%⁷, and 48.4%⁸ of NHL cases, respectively. The frequency of DLBCL is marginally higher in Asian countries than in European ones. FL ranks second among NHL cases in Western nations; however, its frequency is comparatively low in Asian countries ^2,9. North America and Western Europe have reported relative percentages of FL ranging from 20–33%; however, the relative incidence of FL in Asian countries is approximately 10%^2,9. From 2008–2010, FL accounted for ~ 6–10% of all NHL cases in China¹⁰. From 2009–2015, it accounted for only 7.2% of NHL cases in Southern India¹¹. It is unclear whether this variation in incidence is linked to the impacts of environmental versus genetic effects or is due to other factors.

Obesity has emerged as a global health issue associated with changes in fat mass (FM) and fat-free mass (FFM), thus resulting in altered energy expenditure, food intake, and appetite regulation¹². The high prevalence of obesity-related diseases worldwide necessitates a comprehensive understanding of the body composition management. The overall trend of obesity is rising; however, heavily populated Asian nations, such as China and India, have comparatively minimal rates of overweight and obese individuals compared to those in developed countries, such as the United Kingdom¹³. In epidemiological research, the characterisation of overweight and obese individuals generally relies on the body mass index (BMI), which has become the most widely used anthropometric indicator (AI) of human health and disease risk^14,15. However, BMI is an inadequate indicator because it does not differentiate body mass composition¹⁶. It should be noted that BMI does not account for racial variations. For example, Asian individuals generally possess a greater amount of body fat mass than Caucasians with the same BMI¹⁴. The association between BMI and cancer risk remains controversial. Observational studies have shown that high BMI may either increase the risk of cancer^17–19, may not be associated with cancer ^20,21, or may even decrease the risk of cancer^22–24. The association between BMI and the risk of NHL has also been inconsistent across studies. Prospective studies by the American Cancer Society²⁵ and the Cancer Epidemiology Group at the University of Oxford²⁶ found that a higher BMI was associated with a higher risk of NHL. However, observational studies in the United States²⁷ and South Korea²⁸ found that obesity did not increase the risk of developing any type of lymphoma. Thus, the relationship between body composition and BMI and NHL remains controversial. These contradictory results suggest that obesity-related metabolic factors or specific body composition may be more relevant risk factors for certain cancers than obesity itself²⁹.

Mendelian randomisation (MR) is a robust tool that derives causal inferences between exposures and outcomes using single nucleotide polymorphisms (SNPs) to infer the causality of an observed association between a modifiable exposure and clinically relevant outcome³⁰. As SNPs are not affected by environmental factors or preclinical diseases, MR estimates are not prone to bias, thereby overcoming some of the disadvantages of traditional epidemiological studies³¹. Moreover, genetic variations occur prior to the disease development; therefore, the results are not affected by reverse causality³⁰. MR has been widely used in the field of epidemiology, and its reliability has been confirmed in a number of studies. For example, a study using MR to explore body composition and the risk of atrial fibrillation found that higher FFM and FM were associated with a higher risk of atrial fibrillation³². In the study by Kleinstern G et al., MR analysis found no association between lipid trait variation and the risk of developing any NHL subtype³³. Li Z et al. performed an MR study to investigate the relationship between COVID-19 and cancer risk, which revealed that COVID-19 was not associated with lymphoma³⁴.

The reason for the higher prevalence of FL and comparable incidence of DLBCL in European countries than in their Asian counterparts requires investigation. Therefore, we focused on DLBCL and FL, which have the highest incidences among NHL subtypes, when exploring AIs to explain the differences in their regional incidence rates. Given the reliability of MR, we aimed to use MR analysis to explore the genetic associations and possible mediators between AIs and NHL. We hypothesised that AIs may be associated with the incidence of DLBCL and FL. Our findings may provide concrete evidence for the development of appropriate public health strategies for NHL.

2.1 Study design

MR analysis is used to investigate the relationship between exposure factors and disease. First, we evaluated the causal relationships between AIs and the incidence of DLBCL and FL, using a two-sample MR approach. Second, to determine whether DLBCL and FL interfere with AIs, we performed a reverse MR analysis. Finally, as basal metabolic rate (BMR) may be a risk factor for some cancers^29,35,36 that are influenced by certain AIs, such as body composition^36,37 and height³⁸, and significant differences exist in terms of BMRs between those of European and those of Asian descent^39–41, we believed this factor might be a potential mediator. To test our study hypothesis, we performed a two-step MR analysis. An overview of the study design is presented in Fig. 1.

2.2 Quality control of SNPs

First, all SNPs were screened based on the following criteria. 1) Relevance: SNPs needed to exhibit strong levels of relevance to outcomes. 2) Independence: SNPs should not be associated with confounding factors related to outcomes (excluding mediators). 3) Exclusivity: SNPs could only influence an outcome through exposure, and therefore, should not be directly linked to the outcome. Based on these screening principles, we first assessed SNPs with p values of < 5 × 10^− 8. If the number of SNPs was < 20, we used a less strict genome-wide significance threshold of p < 5 × 10^− 642. Second, we established r² < 0.001 and kb = 10,000 as thresholds to exclude linkage imbalances⁴³. Third, to eliminate potential bias from weak instrumental variables (IVs), SNPs with weak statistical power (i.e., F-statistic < 10) were also excluded⁴⁴. Finally, palindromically incompatible SNPs and SNPs detected using the Steiger Filter were removed⁴⁵. A schematic of the three hypotheses of our MR study is shown in Fig. 2.

There are three assumptions of MR design. 1) Relevance - SNPs must exhibit a strong relevance with outcomes. 2) Independence - SNPs must not be associated with confounding factors related to outcomes (excluding mediators). 3) Exclusivity - SNPs can only influence an outcome through exposure and cannot be directly linked to an outcome. In the diagram, the dotted line indicates the absence of causation, while the continuous line signifies its presence.

2.3 Data sources

The integrated epidemiology unit (IEU) database (https://gwas.mrcieu.ac.uk/), which is supported by the MRC IEU at the University of Bristol, has collated and published genome-wide association study (GWAS) data from the UKB and FinnGen Project databases until 2021, forming the largest publicly available GWAS database aggregation site to-date⁴⁶. The UKB is the largest known UK-based repository of data on genetic and environmental causes of diseases as well as disease prevention. It comprises data collected between 2006 and 2010 from 500,000 UK participants aged 40–69. We considered the following AIs from the UKB as exposures: BMI (sample size = 454,137), whole-body fat mass (WBFM; sample size = 461,950), whole-body fat-free mass (WBFFM; sample size = 454,850), waist circumference (WC; sample size = 462,166), hip circumference (HC; sample size = 462,117), standing height (SH; sample size = 461,950), and weight (Wt; sample size = 461,632). Data on the BMR mediator (sample size = 331,307) were also obtained from the UKB. The FinnGen project is a large genetic research program that aims to explore the relationship between genomic information and health characteristics in the Finnish population⁴⁷. FL and DLBCL were the primary outcomes of this study. GWAS data regarding FL (case size = 522, median age = 63.94, and control size = 180,756) and DLBCL (case size = 209, median age = 67.92, and control size = 174,006) were procured from the FinnGen project and consolidated in the IEU database. Both diagnoses conformed to the ICD10 coding system. All of the GWAS datasets used in this study are available to the public and can be downloaded from IEU database. Further details on the GWAS datasets used in this study are presented in Table 1.

2.4 Confounders

There are a number of potential risk factors associated with NHL. In addition to non-modifiable risk factors, such as genetic predisposition and familial history of haematological malignancies^48,49, studies have also indicated that exposure to certain environmental factors (such as hair dyes⁵⁰, radiation⁵¹, and infection by certain viruses [such as hepatitis B {HBV}⁵², hepatitis C {HCV}⁵³, Epstein-Barr virus {EBV}⁵⁴, and human immunodeficiency virus {HIV}⁵⁵]) as well as Helicobacter pylori (HP) infection⁵⁶ and tobacco smoking⁵⁷ may be associated with an increased risk of NHL. To eliminate possible confounding factors affecting the relationship between our chosen AIs and NHL, we omitted the SNPs known to be linked to potential confounders, including hair dye application; radiation exposure; HBV, HCV, EBV, HIV, and HP infection; and smoking status.

2.5 Statistical analyses

MR was performed using the “TwoSampleMR” (version 7.10.0) in R version 4.3.0(R, Vienna, Austria). For two-sample MR in the absence of horizontal pleiotropy, inverse variance weighting (IVW) was predominantly used⁵⁸, and the primary outcomes were the results obtained from the IVW analysis⁵⁹. The robustness levels of the IVW estimates were assessed using sensitivity analyses, such as MR-Egger (if > 50% of the IVs detect horizontal pleiotropy, the test result is dominant)⁶⁰, weighted-median (if < 50% of the IVs detect horizontal pleiotropy, the test result is dominant)⁶¹, and weighted model methods⁶². The MR-PRESSO method was used to detect and correct for any possible horizontal pleiotropy⁶³. Visualization of sensitivity analysis results were performed by scatter plots, and funnel plots. The results are presented as odds ratios (ORs) and corresponding 95% confidence intervals (95% CIs) for each standard deviation. Standard deviations and 95% CIs were calculated using the delta method⁶⁴.

Finally, we calculated the proportion of each AI’s effect on FL when mediated by BMR, using a two-step MR analysis. Two-step MR analysis was performed on exposures and outcomes, mediators and outcomes, and exposures and mediators, and the β values of their corresponding IVW analyses were recorded as β1, β2, and α, respectively. For our two-step MR analysis, the causal effect of exposure on the outcome (including potential effects via mediators) was the total effect (TE) of the exposure on the outcome. TE consists of direct effect (DE) and indirect effect (IE), where DE represents the casual effect of the exposure on an outcome with an effect value of β1. IE is the mediated effect of an exposure on the outcome, with an effect value of α × β2. The proportion of the mediating effect (ME) was calculated as IE/TE (%). Finally, an interactive mediation test was performed on the value of the mediation effect, with p < 0.05 indicating that the mediation effect existed⁶⁵.

3.1. Quality control of IVs

The statistical performances of all the IVs were high (F-statistic > 10) and passed the MR Steiger filtering step. After MR-PRESSO testing, no outliers were found for any of the SNPs. In our sensitivity analysis, Cochran’s Q test results (Table 2) did not suggest any evidence of heterogeneity (p > 0.05)⁶⁶. An MR-Egger intercept test did not detect directional pleiotropy (p > 0.05)⁶⁷. The MR-PRESSO global test also did not detect horizontal pleiotropy (global test, p > 0.05)⁶⁸, and any outliers detected by MR-PRESSO were eliminated. The results of our directional and horizontal pleiotropy tests are presented in Tables 3 and 4. All palindromic and incompatible SNPs were removed. SNPs related to hair dye application; radiation; HBV, HCV, EBV, HIV, and HP infection; and smoking were also excluded. Detailed information on the excluded SNPs is provided in Table 5. Using our screening criteria for BMI, WBFM, WBFFM, WC, HC, SH, and Wt, we obtained 388, 378, 493, 323, 375, 705, and 50 eligible SNPs, respectively, as IVs to assess their associations with FL. Similarly, 23, 378, 493, 323, 375, 705, and 50 SNPs were screened for DLBCL. The full list of SNPs instrumenting the seven AIs for DLBCL and FL can be found in Table 6.

3.2. Bidirectional two-sample MR analysis

3.2.1. Causal effects of the AIs on DLBCL and FL

For FL, two-sample MR revealed that WBFFM (OR = 1.767, 95% CI = 1.076–2.903, p = 0.024) and SH (OR = 1.379, 95% CI = 1.013–1.876, p = 0.041) were significantly associated with increased risk. It was unclear whether BMI (OR = 1.037, 95% CI = 0.690–1.560; p = 0.860), WBFM (OR = 1.179, 95% CI = 0.770–1.805; p = 0.450), WC (OR = 0.947, 95% CI = 0.570–1.573; p = 0.834), HC (OR = 1.093, 95% CI = 0.725–1.646; p = 0.671), and Wt (OR = 1.093, 95% CI = 0.725–1.646; p = 0.671) had causal effects on FL. Scatter and funnel plots showing SNPs linked to the risk of WBFFM, SH, and FL are presented in Figs. 3 and 4. No evidence was found to suggest that BMI (OR = 1.037, 95% CI = 0.690–1.560; p = 0.860), WBFM (OR = 1.037, 95% CI = 0.690–1.560; p = 0.860), WBFFM (OR = 1.037, 95% CI = 0.690–1.560; p = 0.860), WC (OR = 1.037, 95% CI = 0.690–1.560; p = 0.860), HC (OR = 1.037, 95% CI = 0.690–1.560; p = 0.860), SH (OR = 1.037, 95% CI = 0.690–1.560; p = 0.860), and Wt (OR = 1.037, 95% CI = 0.690–1.560; p = 0.860) were associated with the development of DLBCL. Figure 5 presents the results of the two-sample MR analysis using IVW.

Figures 3A and 3B show scatter plots presenting SNPs linked to the risk of WHFFM, SH and the onset of FL. Results are shown for different MR methods, including inverse variance weighted, weighted median, MR-Egger and weighted mode. The plot displays the effect sizes of SNP-positive exposure associations (presented on the X-axis in SD units) and SNP-FL associations (presented on the Y-axis in log _(OR)) with a 95% confidence interval.

Figures 4A and 4B show funnel plots of SNPs associated with the risk of positive exposure (WHFFM, SH) and FL incidence, which is used to evaluate robustness. The scattered dots representing the effects estimated using a single SNP as an instrumental variable. The vertical line represents the population estimate obtained by the variance-weighted inverse estimate and the MR-Egger regression.

3.2.2 Reverse causal effects of FL and DLBCL on AIs

We performed a bidirectional MR analysis to determine whether there were any reverse causal effects between the AIs and DLBCL/FL. The results revealed that genetic predisposition to DLBCL and FL did not influence any of the AIs. Figure 6 presents the results of this reverse two-sample MR analysis using IVW.

The IVW test results of MR analysis of Two-sample MR analysis, where positive causality is bolded.

The IVW test results of MR analysis of Two-sample MR analysis. Abbreviations: AIs, anthropometric indicators; FL, follicular lymphoma; DLBCL, diffuse large B-cell lymphoma; BMR, basal metabolic rate; MR, Mendelian randomization; GWAS, Genome-wide Association Study; BMI, body mass index; WBFM, whole body fat mass; WBFFM, whole body fat-free mass; WC, waist circumference; HC, hip circumference; SH, standing height; Wt, weight; DE, direct effect; IE, indirect effect; TE, total effect; ME, mediating effect; UKB, UK Biobank; SNPs, single nucleotide polymorphisms; IVs, instrumental variables; IVW, Inverse Variance Weighting; NO, number; OR, odds rations; CI, confidence interval; p, p-value.

3.3 Two-step MR analysis

The DE values (β1) of WBFFM (β1_WBFFM = 0.244) and SH (β1_SH = 0.156) for the outcome were obtained using two-sample MR. Owing to the lack of clear causal connections between BMI, WBFM, WC, and HC and FL or DLBCL, mediation analysis was performed using WBFFM and SH only. A positive correlation was found between WBFFM (nSNP = 21; OR = 2.308, 95% CI = 2.097–2.541; p = 2.627 × 10^− 65; α_WBFFM = 0.837) and SH (nSNP = 335; OR = 1.368, 95% CI = 1.340–1.397; p = 3.005 × 10^− 188; α_SH = 0.313) and BMR. Moreover, an increase in BMR (1 SD = 1,357.830 KJ) was found to increase the risk of developing FL (OR = 1.792, 95% CI = 1.092–2.940; p = 0.021; β2 = 0.583). The IE values of WBFFM and SH for FL (mediated by BMR) were 0.487 and 0.182, respectively. Regarding the causal effects of WBFFM and SH on FL, the MF proportions of the BMR were 66.617% (p = 0.022) and 53.909% (p = 0.021), respectively. The detailed results of this two-step MR analysis are presented in Table 7.

Using a two-sample MR study, we identified two key findings. First, for every 1 SD increase in WBFFM (1 SD = 11.449 kg) and SH (1 SD = 9.276 cm), the likelihood of FL occurrence increased by 1.76- and 1.379-fold, respectively. Second, there was no evidence of a causal association between any of the AIs we studied and DLBCL. Our reverse MR analysis revealed that none of the causal relationships exhibited bidirectionality. Finally, our two-step MR analysis found that the causal relationships between WBFFM and SH and FL incidence were mediated by BMR, with mediated proportions of 69.599% and 54.401%, respectively.

Studies have reported that for the same BMI, individuals of Asian descent have higher body FMs, whereas those of European descent have higher FFMs^69,70. FFM, or lean body mass, consists mainly of skeletal muscle and is vital for both mobility and metabolic functions⁷¹. It is believed that FFM is a highly heritable characteristic, based on studies in families and twins⁷², with an estimated heritability rate of 0.52-0.60^73,74. Previous observational^75–77 and MR studies⁷⁸ have investigated the association between FFM and the risk of developing NHL. For example, the MR analysis performed by Vithayathil M et al. revealed that NHL risk increased by 1.89-fold (95% CI = 1.21–2.96; p = 0.005) for every 18.6 ± 2.6 kg/m² increase in FFM index⁷⁸. However, none of these studies addressed which NHL types are particularly affected by FFM. Our study demonstrated that higher the WBFFM higher the risk of FL, thereby supporting the findings of previous MR-based⁷⁸ and observational studies⁷⁵ on the association between WBFFM and NHL. Furthermore, it is plausible that the increased risk of developing FL in the European population is associated with relatively high FFM.

A number of studies have reported that Europeans have been taller than Asians over the last 100 years⁷⁹, and a link between height and the development of NHL has been demonstrated in both adolescents^75,76 and adults⁸⁰. For instance, Bertrand KA et al.⁷⁵ found a positive correlation between height and the onset of NHL in adolescents in their observational study (hazard ratio = 1.004, 95% CI = 1.002–1.005; p = 4.29 × 10^− 9). A cohort study by Murphy F et al.⁸¹ also found that the risk of FL increased by 1.28-fold for every 10 centimeter increase in height. Our study revealed a correlation between increased height and increased susceptibility to FL in adults, which suggested that the differential incidence of FL among European and Asian populations may be linked to differences in height. However, our findings contradict those of the MR study performed by Moore A et al.⁸², who did not find any association between height and FL prevalence. We hypothesise that two factors contributed to this difference. First, as both outcomes and exposures in Moore A et al.'s study pertained to European populations, there may have been some overlapping exposure and outcome samples. By contrast, in this study, the outcome GWAS data were sourced from the UKB database and the exposure GWAS data originated from the FinnGen project, thereby avoiding the possibility of sample overlap. Second, as our study comprised a smaller number of cases than the study by Moore A et al., we cannot dismiss the possibility that our findings are less conclusive. This implies the need for a larger sample to further investigate the correlation between SH and the development of adult-onset FL.

Our MR study found no evidence of causal associations between BMI, WBFM, WC, and HC and FL, consistent with the findings of Vithayathil M et al.⁷⁸. We also found that the onset of DLBCL was not associated with any of the AIs. Previous studies have reported no significant correlations between BMI⁸³, height⁸², WC⁸⁴, HC⁸⁵, Wt⁸⁶, and even obesity⁸⁵ the onset of DLBCL. One reason for the comparable incidence of DLBCL in European and Asian populations may be that AIs do not have an impact on DLBCL incidence. Contrary to our findings, some observational studies suggested that higher BMI in early adulthood is associated with a higher risk of NHL^75,87. For example, one study on Japanese individuals showed that obesity during early adulthood (around 20 years of age) was linked to an increased risk of NHL, particularly DLBCL⁸⁷. The differences in our results may also be due to the median age of the study population, which was 69.96 years.

Although studies have identified a correlation between FFM and NHL, there is a lack of studies on the potential mediating mechanisms behind this relationship. Since the early 20th century, studies have noted that the BMR of Asian populations is lower than that of their American and European counterparts³⁹. This notion has been supported by a number of studies ^40,41. There is strong evidence that FFM, rather than FM, is the main determinant of BMR⁸⁸. Individuals with high FFM and tall statures usually tailor their eating behaviours to meet the energy needs of their tissues and organs, and these mechanisms are mediated by BMR^38,89. However, limited studies have been performed on the potential risks associated with BMR and FL. However, many studies have reported an increased risk of other cancers, including breast cancer, pancreatic adenocarcinoma, postmenopausal breast cancer, and endometrial cancer, associated with a high BMR, even in normal-weight individuals²⁹. Our two-step MR study presented compelling evidence supporting the notion that high WBFFM and SH increase BMR and thus, FL risk as well. The biological mechanisms that underlie this relationship can be explained in three ways. First, individuals with higher BMRs require more cellular energy production to meet their higher energy and metabolic needs, and increased aerobic glycolysis can lead to the production of large amounts of reactive oxygen species, in turn leading to increased protein, lipid, and DNA damage^90,91. The accumulation of errors in this process puts pressure on the body's repair system and increases the risk of mutations, thus significantly increasing the risk of cancer^29,92. Second, in vivo experimental studies have shown that a high BMR reduces the effectiveness of the immune response and can thus lead to cancer development^93,94. Third, a high BMR can lead to a pro-inflammatory state⁹⁵, which has been observed to be associated with a higher risk of developing NHL⁸⁶. In a European case-control study, elevated levels of inflammatory biomolecules (IL-2, ICAM, IFN-γ, and TNF-α) were found to be associated with the risk of developing NHL⁹⁶.

This study has some limitations. First, to minimise statistical bias due to demographic stratification, the population we studied was entirely European; therefore, our findings may not be applicable to other ethnic groups. Second, our study aimed to analyse specific subtypes of lymphoma, and the FL and DLBCL GWAS data included a small number of cases, which may have led to some false-positive results. Third, our MR study evaluated the risk of FL using genetic IVs of whole-body FFM and SH, but it did not consider that the relationship between body composition and FL risk may change over time (although we did implement robust methods to assess the sensitivity and robustness of this hypothesis). Fourth, while certain studies have indicated that BMR is not significantly affected by sex ⁹⁷, there are notable discrepancies in terms of WBFFM^98,99 and SH¹⁰⁰ between males and females. As the GWAS data on FL used for this study did not differentiate between the sexes, we could not examine the stratified risk of developing FL in men vs women. Thus, future MR-based research stratified by sex is warranted in this field. Finally, our two-step MR study revealed that BMR had MEs of 66•617% and 53•909% on WBFFM and SH, respectively, thereby indicating the likely presence of other unknown mediating factors.

We observed a causal relationship between WBFFM and SH and the onset of FL in Europeans. However, we did not observe any causal association between AIs and DLBCL. Our findings may aid in understanding the differences that have been noted in the relative prevalence of FL and DLBCL between European and Asian populations. The finding may help design strategies to prevent cancer development. In the future, further studies are worthwhile to explore the correlation between WBFFM and SH and the onset of FL in the pathogenesis and treatment strategies.

Acknowledgment

We thank the UK Biobank for providing GWAS pooled data on anthropometric indicators, and the participants and investigators of the FinnGen study for providing GWAS pooled data on diffuse large B-cell lymphoma and follicular lymphoma to this study.

Author contributions

Y.Z., L.H., and Z.L. designed the study; X.Z., G.Z., X.L. and Y.D. collected the data; Y.Z., X.Z. and T.G. analysed the data; Y.Z., X.Z., G.Z. interpreted the data; All authors drafted the manuscript; Y.Z., Z.L., G.Z., L.H., X.Z. approved the final version of the manuscript.

Data availability statement

Any data access request proposals should be directed to [email protected] for the consideration of the study's prin-cipal investigators.

Declaration of interests

The authors declare no competing interests.

Ethical approval

This study was based on the data from a public UK Biobank (UKB; http://www.ukbiobank.ac.uk/) and the FinnGen (https://r5.finngen.fi/) database. Ethical approval and consent to participate were obtained during the original data collection; hence, the requirement for further ethical approval was waived for this study.

Shankland, K. R., Armitage, J. O. & Hancock, B. W. Non-Hodgkin lymphoma. Lancet (London, England) 380, 848–857, doi:10.1016/s0140-6736(12)60605-9 (2012).
Thandra, K. C. et al. Epidemiology of Non-Hodgkin's Lymphoma. Medical sciences (Basel, Switzerland) 9, doi:10.3390/medsci9010005 (2021).
Miret, M. et al. Incidence of second primary malignancies in relapsed/refractory B-cell non-Hodgkin's lymphoma patients in England. Leukemia research 127, 107042, doi:10.1016/j.leukres.2023.107042 (2023).
Apostolidis, J., Mokhtar, N., Al Omari, R., Darweesh, M. & Al Hashmi, H. Follicular lymphoma: Update on management and emerging therapies at the dawn of the new decade. Hematological oncology 38, 213–222, doi:10.1002/hon.2711 (2020).
Susanibar-Adaniya, S. & Barta, S. K. 2021 Update on Diffuse large B cell lymphoma: A review of current data and potential applications on risk stratification and management. American journal of hematology 96, 617–629, doi:10.1002/ajh.26151 (2021).
Xiao-Qiu, L. I. et al. Distribution pattern of lymphoma subtypes in China: A nationwide multicenter study of 10002 cases. (2012).
Muto, R. et al. Epidemiology and secular trends of malignant lymphoma in Japan: Analysis of 9426 cases according to the World Health Organization classification. Cancer medicine 7, 5843–5858, doi:10.1002/cam4.1805 (2018).
Kim, J. S. et al. Increasing Incidence of B-Cell Non-Hodgkin Lymphoma and Occurrence of Second Primary Malignancies in South Korea: 10-Year Follow-up Using the Korean National Health Information Database. Cancer research and treatment 52, 1262–1272, doi:10.4143/crt.2020.089 (2020).
Cho, S. H. et al. Clinical Features and Survival of Patients With Follicular Lymphoma in Korea. Clinical lymphoma, myeloma & leukemia 16, 197–202, doi:10.1016/j.clml.2015.12.012 (2016).
Yang, Q. P. et al. Subtype distribution of lymphomas in Southwest China: analysis of 6,382 cases using WHO classification in a single institution. Diagnostic pathology 6, 77, doi:10.1186/1746-1596-6-77 (2011).
Sylvia, M. T. et al. Follicular Lymphoma: A Clinicopathological Analysis from a Tertiary Care Institute in Southern India. Mediterranean journal of hematology and infectious diseases 8, e2016060, doi:10.4084/mjhid.2016.060 (2016).
Simonson, M., Boirie, Y. & Guillet, C. Protein, amino acids and obesity treatment. Reviews in endocrine & metabolic disorders 21, 341–353, doi:10.1007/s11154-020-09574-5 (2020).
Ramachandran, A., Chamukuttan, S., Shetty, S. A., Arun, N. & Susairaj, P. Obesity in Asia–is it different from rest of the world. Diabetes/metabolism research and reviews 28 Suppl 2, 47–51, doi:10.1002/dmrr.2353 (2012).
Chooi, Y. C., Ding, C. & Magkos, F. The epidemiology of obesity. Metabolism: clinical and experimental 92, 6–10, doi:10.1016/j.metabol.2018.09.005 (2019).
Yang, H. I. et al. A new anthropometric index to predict percent body fat in young adults. Public health nutrition 23, 1507–1514, doi:10.1017/s1368980019004191 (2020).
Merchant, R. A. et al. Relationship of Fat Mass Index and Fat Free Mass Index With Body Mass Index and Association With Function, Cognition and Sarcopenia in Pre-Frail Older Adults. Frontiers in endocrinology 12, 765415, doi:10.3389/fendo.2021.765415 (2021).
Zheng, R. et al. Body mass index (BMI) trajectories and risk of colorectal cancer in the PLCO cohort. British journal of cancer 119, 130–132, doi:10.1038/s41416-018-0121-y (2018).
Sohn, W. et al. Obesity and the risk of primary liver cancer: A systematic review and meta-analysis. Clinical and molecular hepatology 27, 157–174, doi:10.3350/cmh.2020.0176 (2021).
Chen, J. et al. Body Mass Index and Cancer Risk: An Umbrella Review of Meta-Analyses of Observational Studies. Nutrition and cancer 75, 1051–1064, doi:10.1080/01635581.2023.2180824 (2023).
Song, X. et al. Body mass index and cancer incidence: the FINRISK study. European journal of epidemiology 29, 477–487, doi:10.1007/s10654-014-9934-z (2014).
Renehan, A. G., Tyson, M., Egger, M., Heller, R. F. & Zwahlen, M. Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet (London, England) 371, 569–578, doi:10.1016/s0140-6736(08)60269-x (2008).
Zhou, W. et al. Causal relationships between body mass index, smoking and lung cancer: Univariable and multivariable Mendelian randomization. International journal of cancer 148, 1077–1086, doi:10.1002/ijc.33292 (2021).
You, D. et al. Associations of genetic risk, BMI trajectories, and the risk of non-small cell lung cancer: a population-based cohort study. BMC medicine 20, 203, doi:10.1186/s12916-022-02400-6 (2022).
Shao, C. et al. Body mass index and glioma risk: A prospective multicenter study. Frontiers in endocrinology 13, 933921, doi:10.3389/fendo.2022.933921 (2022).
Calle, E. E., Rodriguez, C., Walker-Thurmond, K. & Thun, M. J. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. The New England journal of medicine 348, 1625–1638, doi:10.1056/NEJMoa021423 (2003).
Reeves, G. K. et al. Cancer incidence and mortality in relation to body mass index in the Million Women Study: cohort study. BMJ (Clinical research ed.) 335, 1134, doi:10.1136/bmj.39367.495995.AE (2007).
Samanic, C. et al. Obesity and cancer risk among white and black United States veterans. Cancer causes & control: CCC 15, 35–43, doi:10.1023/B:CACO.0000016573.79453.ba (2004).
Oh, S. W., Yoon, Y. S. & Shin, S. A. Effects of excess weight on cancer incidences depending on cancer sites and histologic findings among men: Korea National Health Insurance Corporation Study. Journal of clinical oncology: official journal of the American Society of Clinical Oncology 23, 4742–4754, doi:10.1200/jco.2005.11.726 (2005).
Kliemann, N. et al. Predicted basal metabolic rate and cancer risk in the European Prospective Investigation into Cancer and Nutrition. International journal of cancer 147, 648–661, doi:10.1002/ijc.32753 (2020).
Smith, G. D. & Ebrahim, S. 'Mendelian randomization': can genetic epidemiology contribute to understanding environmental determinants of disease? Int J Epidemiol 32, 1–22, doi:10.1093/ije/dyg070 (2003).
Bowden, J. & Holmes, M. V. Meta-analysis and Mendelian randomization: A review. Res Synth Methods 10, 486–496, doi:10.1002/jrsm.1346 (2019).
Tikkanen, E. et al. Body composition and atrial fibrillation: a Mendelian randomization study. European heart journal 40, 1277–1282, doi:10.1093/eurheartj/ehz003 (2019).
Kleinstern, G. et al. Lipid Trait Variants and the Risk of Non-Hodgkin Lymphoma Subtypes: A Mendelian Randomization Study. Cancer epidemiology, biomarkers & prevention: a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology 29, 1074–1078, doi:10.1158/1055-9965.Epi-19-0803 (2020).
Li, Z., Wei, Y., Zhu, G., Wang, M. & Zhang, L. Cancers and COVID-19 Risk: A Mendelian Randomization Study. Cancers 14, doi:10.3390/cancers14092086 (2022).
Ng, J. C. M. & Schooling, C. M. Effect of Basal Metabolic Rate on Cancer: A Mendelian Randomization Study. Frontiers in genetics 12, 735541, doi:10.3389/fgene.2021.735541 (2021).
Cornish, A. J. et al. Modifiable pathways for colorectal cancer: a mendelian randomisation analysis. The lancet. Gastroenterology & hepatology 5, 55–62, doi:10.1016/s2468-1253(19)30294-8 (2020).
Zhang, Y. et al. [Basal metabolic rate of overweight and obese adults in Beijing]. Wei sheng yan jiu = Journal of hygiene research 45, 739–748 (2016).
Zou, Y., Wang, Q. & Cheng, X. Causal Relationship Between Basal Metabolic Rate and Alzheimer's Disease: A Bidirectional Two-sample Mendelian Randomization Study. Neurology and therapy 12, 763–776, doi:10.1007/s40120-023-00458-9 (2023).
de Boer, J. O., van Es, A. J., Voorrips, L. E., Blokstra, F. & Vogt, J. E. Energy metabolism and requirements in different ethnic groups. European journal of clinical nutrition 42, 983–997 (1988).
Ismail, M. N. et al. Obesity in Malaysia. Obesity reviews: an official journal of the International Association for the Study of Obesity 3, 203–208, doi:10.1046/j.1467-789x.2002.00074.x (2002).
Vyas, A. et al. Nutrient intakes of an adult Pakistani, European and African-Caribbean community in inner city Britain. Journal of human nutrition and dietetics: the official journal of the British Dietetic Association 16, 327–337, doi:10.1046/j.1365-277x.2003.00461.x (2003).
Chen, L. et al. Examination on the risk factors of cholangiocarcinoma: A Mendelian randomization study. Frontiers in pharmacology 13, 900424, doi:10.3389/fphar.2022.900424 (2022).
Hemani, G., Tilling, K. & Davey Smith, G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS genetics 13, e1007081, doi:10.1371/journal.pgen.1007081 (2017).
Burgess, S. & Thompson, S. G. Avoiding bias from weak instruments in Mendelian randomization studies. International journal of epidemiology 40, 755–764, doi:10.1093/ije/dyr036 (2011).
Wu, E., Ni, J., Tao, L. & Xie, T. A bidirectional Mendelian randomization study supports the causal effects of a high basal metabolic rate on colorectal cancer risk. PloS one 17, e0273452, doi:10.1371/journal.pone.0273452 (2022).
Yang, W. et al. Dietary factors and risk for asthma: A Mendelian randomization analysis. Frontiers in immunology 14, 1126457, doi:10.3389/fimmu.2023.1126457 (2023).
Kurki, M. I. et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature 613, 508–518, doi:10.1038/s41586-022-05473-8 (2023).
Chang, E. T. et al. Family history of hematopoietic malignancy and risk of lymphoma. Journal of the National Cancer Institute 97, 1466–1474, doi:10.1093/jnci/dji293 (2005).
Linet, M. S. et al. Medical history, lifestyle, family history, and occupational risk factors for follicular lymphoma: the InterLymph Non-Hodgkin Lymphoma Subtypes Project. Journal of the National Cancer Institute. Monographs 2014, 26–40, doi:10.1093/jncimonographs/lgu006 (2014).
Zhang, Y. et al. Personal use of hair dye and the risk of certain subtypes of non-Hodgkin lymphoma. American journal of epidemiology 167, 1321–1331, doi:10.1093/aje/kwn058 (2008).
Kim, C. J., Freedman, D. M., Curtis, R. E., Berrington de Gonzalez, A. & Morton, L. M. Risk of non-Hodgkin lymphoma after radiotherapy for solid cancers. Leukemia & lymphoma 54, 1691–1697, doi:10.3109/10428194.2012.753543 (2013).
Bouvard, V. et al. A review of human carcinogens–Part B: biological agents. The Lancet. Oncology 10, 321–322, doi:10.1016/s1470-2045(09)70096-8 (2009).
Dal Maso, L. & Franceschi, S. Hepatitis C virus and risk of lymphoma and other lymphoid neoplasms: a meta-analysis of epidemiologic studies. Cancer epidemiology, biomarkers & prevention: a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology 15, 2078–2085, doi:10.1158/1055-9965.Epi-06-0308 (2006).
Mackrides, N. et al. Epstein-Barr virus-positive follicular lymphoma. Modern pathology: an official journal of the United States and Canadian Academy of Pathology, Inc 30, 519–529, doi:10.1038/modpathol.2016.214 (2017).
Bohlius, J. et al. Incidence and risk factors of HIV-related non-Hodgkin's lymphoma in the era of combination antiretroviral therapy: a European multicohort study. Antiviral therapy 14, 1065–1074, doi:10.3851/imp1462 (2009).
Biological agents. IARC monographs on the evaluation of carcinogenic risks to humans 100, 1-441 (2012).
Lu, Y. et al. Cigarette smoking, passive smoking, and non-Hodgkin lymphoma risk: evidence from the California Teachers Study. American journal of epidemiology 174, 563–573, doi:10.1093/aje/kwr127 (2011).
Burgess, S., Butterworth, A. & Thompson, S. G. Mendelian randomization analysis with multiple genetic variants using summarized data. Genetic epidemiology 37, 658–665, doi:10.1002/gepi.21758 (2013).
Bowden, J. et al. Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: the role of the I2 statistic. Int J Epidemiol 45, 1961–1974, doi:10.1093/ije/dyw220 (2016).
Burgess, S., Bowden, J., Fall, T., Ingelsson, E. & Thompson, S. G. Sensitivity Analyses for Robust Causal Inference from Mendelian Randomization Analyses with Multiple Genetic Variants. Epidemiology (Cambridge, Mass.) 28, 30–42, doi:10.1097/ede.0000000000000559 (2017).
Bowden, J., Davey Smith, G., Haycock, P. C. & Burgess, S. Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genetic epidemiology 40, 304–314, doi:10.1002/gepi.21965 (2016).
Hartwig, F. P., Davey Smith, G. & Bowden, J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. International journal of epidemiology 46, 1985–1998, doi:10.1093/ije/dyx102 (2017).
Abuduxukuer, R., Niu, P. P., Guo, Z. N., Xu, Y. M. & Yang, Y. Circulating Insulin-Like Growth Factor 1 Levels and Migraine Risk: A Mendelian Randomization Study. Neurology and therapy 11, 1677–1689, doi:10.1007/s40120-022-00398-w (2022).
Burgess, S., Daniel, R. M., Butterworth, A. S. & Thompson, S. G. Network Mendelian randomization: using genetic variants as instrumental variables to investigate mediation in causal pathways. Int J Epidemiol 44, 484–495, doi:10.1093/ije/dyu176 (2015).
Sanderson, E. Multivariable Mendelian Randomization and Mediation. Cold Spring Harbor perspectives in medicine 11, doi:10.1101/cshperspect.a038984 (2021).
Li, N. et al. Causal Effects of Specific Gut Microbiota on Chronic Kidney Diseases and Renal Function-A Two-Sample Mendelian Randomization Study. Nutrients 15, doi:10.3390/nu15020360 (2023).
Burgess, S. & Thompson, S. G. Interpreting findings from Mendelian randomization using the MR-Egger method. European journal of epidemiology 32, 377–389, doi:10.1007/s10654-017-0255-x (2017).
Verbanck, M., Chen, C. Y., Neale, B. & Do, R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nature genetics 50, 693–698, doi:10.1038/s41588-018-0099-7 (2018).
Deurenberg, P., Yap, M. & van Staveren, W. A. Body mass index and percent body fat: a meta analysis among different ethnic groups. International journal of obesity and related metabolic disorders: journal of the International Association for the Study of Obesity 22, 1164–1171, doi:10.1038/sj.ijo.0800741 (1998).
Deurenberg, P., Deurenberg-Yap, M. & Guricci, S. Asians are different from Caucasians and from each other in their body mass index/body fat per cent relationship. Obesity reviews: an official journal of the International Association for the Study of Obesity 3, 141–146, doi:10.1046/j.1467-789x.2002.00065.x (2002).
Sergi, G., De Rui, M., Stubbs, B., Veronese, N. & Manzato, E. Measurement of lean body mass using bioelectrical impedance analysis: a consideration of the pros and cons. Aging clinical and experimental research 29, 591–597, doi:10.1007/s40520-016-0622-6 (2017).
Zillikens, M. C. et al. Large meta-analysis of genome-wide association studies identifies five loci for lean body mass. Nature communications 8, 80, doi:10.1038/s41467-017-00031-7 (2017).
Arden, N. K. & Spector, T. D. Genetic influences on muscle strength, lean body mass, and bone mineral density: a twin study. Journal of bone and mineral research: the official journal of the American Society for Bone and Mineral Research 12, 2076–2081, doi:10.1359/jbmr.1997.12.12.2076 (1997).
Hsu, F. C. et al. Heritability of body composition measured by DXA in the diabetes heart study. Obesity research 13, 312–319, doi:10.1038/oby.2005.42 (2005).
Bertrand, K. A. et al. A prospective analysis of body size during childhood, adolescence, and adulthood and risk of non-Hodgkin lymphoma. Cancer prevention research (Philadelphia, Pa.) 6, 864–873, doi:10.1158/1940-6207.Capr-13-0132 (2013).
Odutola, M. K. et al. Associations between early-life growth pattern and body size and follicular lymphoma risk and survival: a family-based case-control study. Cancer epidemiology 80, 102241, doi:10.1016/j.canep.2022.102241 (2022).
Leiba, M. et al. Adolescent weight and height are predictors of specific non-Hodgkin lymphoma subtypes among a cohort of 2,352,988 individuals aged 16 to 19 years. Cancer 122, 1068–1077, doi:10.1002/cncr.29792 (2016).
Vithayathil, M. et al. Body size and composition and risk of site-specific cancers in the UK Biobank and large international consortia: A mendelian randomisation study. PLoS medicine 18, e1003706, doi:10.1371/journal.pmed.1003706 (2021).
A century of trends in adult human height. eLife 5, doi:10.7554/eLife.13410 (2016).
Morton, L. M. et al. Etiologic heterogeneity among non-Hodgkin lymphoma subtypes: the InterLymph Non-Hodgkin Lymphoma Subtypes Project. Journal of the National Cancer Institute. Monographs 2014, 130–144, doi:10.1093/jncimonographs/lgu013 (2014).
Murphy, F. et al. Body size in relation to incidence of subtypes of haematological malignancy in the prospective Million Women Study. British journal of cancer 108, 2390–2398, doi:10.1038/bjc.2013.159 (2013).
Moore, A. et al. Genetically Determined Height and Risk of Non-hodgkin Lymphoma. Frontiers in oncology 9, 1539, doi:10.3389/fonc.2019.01539 (2019).
Sarkozy, C., Camus, V., Tilly, H., Salles, G. & Jardin, F. Body mass index and other anthropometric parameters in patients with diffuse large B-cell lymphoma: physiopathological significance and predictive value in the immunochemotherapy era. Leukemia & lymphoma 56, 1959–1968, doi:10.3109/10428194.2014.979412 (2015).
Britton, J. A. et al. Anthropometric characteristics and non-Hodgkin's lymphoma and multiple myeloma risk in the European Prospective Investigation into Cancer and Nutrition (EPIC). Haematologica 93, 1666–1677, doi:10.3324/haematol.13078 (2008).
Hidayat, K., Li, H. J. & Shi, B. M. Anthropometric factors and non-Hodgkin's lymphoma risk: systematic review and meta-analysis of prospective studies. Critical reviews in oncology/hematology 129, 113–123, doi:10.1016/j.critrevonc.2018.05.018 (2018).
Willett, E. V. et al. Non-Hodgkin lymphoma and obesity: a pooled analysis from the InterLymph Consortium. International journal of cancer 122, 2062–2070, doi:10.1002/ijc.23344 (2008).
Kanda, J. et al. Association between obesity and the risk of malignant lymphoma in Japanese: a case-control study. International journal of cancer 126, 2416–2425, doi:10.1002/ijc.24955 (2010).
Johannsen, D. L. et al. Metabolic slowing with massive weight loss despite preservation of fat-free mass. The Journal of clinical endocrinology and metabolism 97, 2489–2496, doi:10.1210/jc.2012-1444 (2012).
Hopkins, M., Gibbons, C. & Blundell, J. Fat-free mass and resting metabolic rate are determinants of energy intake: implications for a theory of appetite control. Philosophical transactions of the Royal Society of London. Series B, Biological sciences 378, 20220213, doi:10.1098/rstb.2022.0213 (2023).
Dang, C. V. Links between metabolism and cancer. Genes & development 26, 877–890, doi:10.1101/gad.189365.112 (2012).
Stier, A., Reichert, S., Massemin, S., Bize, P. & Criscuolo, F. Constraint and cost of oxidative stress on reproduction: correlative evidence in laboratory mice and review of the literature. Frontiers in zoology 9, 37, doi:10.1186/1742-9994-9-37 (2012).
Porporato, P. E., Filigheddu, N., Pedro, J. M. B., Kroemer, G. & Galluzzi, L. Mitochondrial metabolism and cancer. Cell research 28, 265–280, doi:10.1038/cr.2017.155 (2018).
Książek, A. & Konarzewski, M. Effect of dietary restriction on immune response of laboratory mice divergently selected for basal metabolic rate. Physiological and biochemical zoology: PBZ 85, 51–61, doi:10.1086/663696 (2012).
Råberg, L. et al. Basal metabolic rate and the evolution of the adaptive immune system. Proceedings. Biological sciences 269, 817–821, doi:10.1098/rspb.2001.1953 (2002).
Ferrucci, L. et al. The origins of age-related proinflammatory state. Blood 105, 2294–2299, doi:10.1182/blood-2004-07-2599 (2005).
Larsson, S. C. & Wolk, A. Obesity and risk of non-Hodgkin's lymphoma: a meta-analysis. International journal of cancer 121, 1564–1570, doi:10.1002/ijc.22762 (2007).
Saberi Hosnijeh, F. et al. Plasma cytokines and future risk of non-Hodgkin lymphoma (NHL): a case-control study nested in the Italian European Prospective Investigation into Cancer and Nutrition. Cancer epidemiology, biomarkers & prevention: a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology 19, 1577–1584, doi:10.1158/1055-9965.Epi-09-1237 (2010).
Lamont, L. S., Romito, R. & Rossi, K. Fat-free mass and gender influences the rapid-phase excess postexercise oxygen consumption. Applied physiology, nutrition, and metabolism = Physiologie appliquee, nutrition et metabolisme 35, 23–26, doi:10.1139/h09-120 (2010).
Lu, Y. et al. Comparison of fat-free mass index and fat mass index in Chinese adults. European journal of clinical nutrition 66, 1004–1007, doi:10.1038/ejcn.2012.86 (2012).
Rosenfeld, R. G. Gender differences in height: an evolutionary perspective. Journal of pediatric endocrinology & metabolism: JPEM 17 Suppl 4, 1267–1271 (2004).

Tables 1-7 is available in the Supplementary Files section.

No competing interests reported.

Table17.xlsx

Download PDF

Journal Publication

published 01 Jun, 2024

Read the published version in Gene →

Version 1

posted

You are reading this latest preprint version

Anthropometric indicators may explain the high incidence of follicular lymphoma in Europeans: Results from a bidirectional two-sample two-step Mendelian randomisation

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Conclusions

Figures

1. Introduction

2. Materials and Methods

2.1 Study design

2.2 Quality control of SNPs

2.3 Data sources

2.4 Confounders

2.5 Statistical analyses

3 Results

3.1. Quality control of IVs

3.2. Bidirectional two-sample MR analysis

3.2.1. Causal effects of the AIs on DLBCL and FL

3.2.2 Reverse causal effects of FL and DLBCL on AIs

3.3 Two-step MR analysis

4 Discussion

5. Conclusions

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1