Incorporating biological and clinical insights into variant choice for Mendelian randomisation: examples and principles

Novel insights into immune-gut microbiota interactions in colorectal cancer: a Mendelian randomization study

BACKGROUND

The relationship between immune cells and colorectal cancer (CRC) development has been extensively studied; however, the mediating role of gut microbiota in this relationship remains poorly understood.

METHODS

We utilized summary data from genome-wide association studies (GWAS) to analyze 731 immune cell phenotypes, 473 gut microbiota, and CRC-related data. A two-step mediation analysis was employed to identify mediating gut microbiota. The primary analysis method was inverse variance weighting (IVW), supplemented by MR-Egger, simple mode, weighted median, and weighted mode analyses. Robustness of the results was ensured through systematic sensitivity analyses.

RESULTS

Our analysis identified 13 immune cell phenotypes significantly associated with CRC, including 10 protective factors and 3 risk factors. Additionally, 13 gut microbiota showed significant associations with CRC, comprising 8 protective factors and 5 risk factors. Mediation analysis revealed that 4-gut microbiota (1 order, 1 family, 1 genus, and 1 unclassified) mediated the relationship between immune cells and CRC. For instance, unclassified CAG - 977 mediated the effects of FSC-A on NK and NKT %lymphocyte on CRC risk, with mediation proportions of 11% and 12.3%, respectively. Notably, 22.3% of the protective effect of EM CD8br %CD8br on CRC was mediated through order Francisellales.

CONCLUSION

This study provides evidence for a potential causal relationship between immune cells, gut microbiota, and CRC, highlighting the mediating role of specific gut microbiota. These findings offer new insights into the pathogenesis of CRC and may inform future therapeutic strategies.

The Lipidomic Profile Discriminates Between MASLD and MetALD

BACKGROUND

The recent consensus statement redefined steatotic liver diseases. Metabolic dysfunction-associated steatotic liver disease (MASLD) and metabolic dysfunction and alcohol-related liver disease (MetALD) now represent distinct disease entities. However, biomarkers that differentiate MASLD and MetALD remain largely unknown.

AIMS

To identify lipidomic biomarkers with discriminatory potential for distinguishing MetALD from MASLD.

METHODS

Using the UK Biobank dataset, 40,534 people with available MRI liver scans were analysed. A total of, 11,217 cases with a proton density fat fraction (PDFF) ≥ 5% were identified as having steatotic liver disease. Among these, lipidomic profiles were obtained for 5539 MASLD and 462 MetALD cases. A total of, 250 plasma lipidomic and metabolomic parameters were analysed. Mendelian randomisation (MR) analysis was used to confirm the association between alcohol consumption and the lipidomic biomarkers.

RESULTS

When comparing the top 30 differentially expressed lipidomic biomarkers predicting MetALD compared to MASLD, the majority were related to HDL and were significantly overrepresented at both analysed time points. The top five metabolites were: acetoacetate, 3-hydroxybutyrate, phospholipids in Large HDL, concentration of large HDL particles, free cholesterol in large HDL. The sensitivity analysis comparing alcohol-related liver disease to MASLD revealed similar associations, suggesting that the HDL signature is stable over time. Additionally, MR analysis further confirmed that alcohol consumption was associated with increased levels of HDL-related metabolites.

CONCLUSION

Our findings indicate that HDL-centric lipidomic markers, particularly those within the larger and medium HDL subfraction, may differentiate MetALD from MASLD. Further longitudinal and experimental studies are warranted to validate these findings and assess their clinical implications.

Examining the link between 179 lipid species and 7 diseases using genetic predictors

BACKGROUND

Genome-wide association studies of lipid species have identified several loci shared with various diseases, however, the relationship between lipid species and disease risk remains poorly understood. Here we investigated whether the plasma levels of lipid species are causally linked to disease risk.

METHODS

We built genetic predictors of 179 lipid species, measured in 7174 Finnish individuals, by utilising either 11 high-impact genomic loci or genome-wide polygenic scores (PGS). We assessed the impact of the lipid species on seven diseases by performing disease association across FinnGen (n = 500,348), UK Biobank (n = 420,531), and Generation Scotland (n = 20,032). We performed univariable Mendelian randomisation (MR) and multivariable MR (MVMR) analyses to examine whether lipid species impact disease risk independently of standard lipids.

FINDINGS

PGS explained >4% of the variance for 34 lipid species but variants outside the high-impact loci had only a marginal contribution. Variants within the high-impact loci showed association with all seven diseases. MVMR supported a causal role of ApoB in ischaemic heart disease after accounting for lipid species. Phosphatidylethanolamine-increasing LIPC variants seemed to lower age-related macular degeneration risk independently of HDL-cholesterol. MVMR suggested a protective effect of four lipid species containing arachidonic acid on cholelithiasis risk independently of Total Cholesterol.

INTERPRETATION

Our study demonstrates how genetic predictors of lipid species can be utilised to gain insights into disease risk. We report potential links between lipid species and age-related macular degeneration and cholelithiasis risk, which can be explored for their utility in disease risk prediction and therapy.

FUNDING

The funders had no role in the study design, data analyses, interpretation, or writing of this article.

Association of free triiodothyronine and total triiodothyronine with nonalcoholic fatty liver disease: from National Health and Nutrition Examination Survey and Mendelian randomization study

BACKGROUND

Nonalcoholic fatty liver disease (NAFLD) is associated with multiple factors, yet the relationship between NAFLD and thyroid-related biomarkers remains unclear. This study aims to elucidate this potential linkage.

METHODS

Utilizing data from the US National Health and Nutrition Examination Survey (NHANES), we explored the possible associations between thyroid-related biomarkers and NAFLD through multivariable logistic regression, subgroup analysis, and interaction tests. A bidirectional Mendelian randomization (MR) approach complemented by various sensitivity analyses was then employed to ascertain these relationships' causality.

RESULTS

Our NHANES analysis indicated significant associations between elevated levels of free triiodothyronine (FT3) [odds ratio (OR): 2.59, 95% confidence interval (CI): 1.50-4.49] and total triiodothyronine (TT3) (OR: 2.01, 95% CI: 1.27-3.18) with the prevalence of NAFLD. MR findings reinforced the causal relationship, demonstrating that NAFLD may elevate FT3 (β: 0.05, 95% CI: 0.01-0.09) and TT3 (β: 0.08, 95% CI: 0.02-0.14) levels. Additionally, thyroid-stimulating hormone (TSH) was confirmed as an independent risk factor for NAFLD (OR: 1.10, 95% CI: 1.04-1.18), with specific MR sensitivity analyses supporting the robustness of these results.

CONCLUSION

This study indicates potential elevations in FT3, TT3, and thyroglobulin levels associated with NAFLD, while also identifying TSH as a risk factor for NAFLD. These findings underscore the importance of routine thyroid function monitoring both in the general population and particularly in individuals with NAFLD.

Association of 91 Inflammatory Factors and 1400 Metabolites with Sepsis: A Mendelian Randomization Analysis

Objective: Observational studies suggest links between inflammatory factors, metabolites, and sepsis, yet their causality is uncertain. This study employs Mendelian Randomization (MR) to investigate the causality between these factors and sepsis, aiming to uncover the precise relationship and identify novel treatment approaches. Methods: We used summary data from genome-wide association studies (GWAS) involving 91 inflammatory factors, 1400 metabolites as exposure, and STREPTO SEPSIS as outcome. Inverse variance weighting (IVW) and MR-Egger were used to evaluate the causal effect between exposure and outcome. Sensitivity analyses were performed using Cochrane's Q test, MR-Egger intercept method, MR-PRESSO method and leave-one-out method. Results: Thymic stromal lymphopoietin levels (TSLP) (OR = 1.269; 95%CI = 1.016-1.585; P = .036) and Interleukin 15 receptor subunit alpha levels (IL-15Rα) (OR = 0.894; 95%CI = 0.801-0.998; P = .046) had a significant causal relationship with sepsis. Forty-four metabolites were associated with sepsis, including Spermidine to choline ratio (OR = 1.447; 95%CI = 1.104-1.977; P = .009), 4-hydroxyhippurate levels (OR = 1.448; 95%CI = 1.117-1.877; P = .005), and Sphingomyelin (d18:1/20:1, d18:2/20:0) levels (OR = 1.371; 95%CI = 1.139-1.651; P < .001). TSLP was associated with 19 metabolites, and IL-15Rα was associated with 30 metabolites. Conclusions: This study uncovers the causal link between sepsis and two inflammatory factors, TSLP and IL-15Rα, and suggests metabolites' potential in intervention. It also identifies 44 metabolites associated with sepsis, indicating possible biomarkers or therapeutic targets. The findings offer new perspectives on sepsis pathogenesis and could inform future treatment strategies.

Unveiling prognostic value of JAK/STAT signaling pathway related genes in colorectal cancer: a study of Mendelian randomization analysis

BACKGROUND

Colorectal cancer (CRC) ranks among the frequently occurring malignant neoplasms affecting the gastrointestinal tract. This study aimed to explore JAK-STAT signaling pathway related genes in CRC and establish a new prognostic model.

METHODS

The data set used in this study is from a public database. JAK-STAT-differentially expressed genes (DEGs) were identified through differential expression analysis and weighted gene co-expression network analysis (WGCNA). Prognostic genes were selected from JAK-STAT-DEGs through Mendelian randomization (MR), univariate Cox regression, and least absolute shrinkage and selection operator (LASSO) analyses. The expressions of prognostic genes were verified by RT-qPCR. Then, a risk model was built and validated by the GSE39582. Independent prognostic factors were screened underlying risk scores and different clinical indicators, resulting in the construction of a nomogram. Additionally, immune infiltration, immune scores and immune checkpoint inhibitors analyses and gene set enrichment analysis (GSEA) were carried out.

RESULTS

The 3,668 JAK-STAT-DEGs were obtained by intersection of 5826 CRC-DEGs and 9766 JAK-STAT key module genes. Five prognostic genes were selected (ANK3, F5, FAM50B, KLHL35, MPP2), and their expressions were significantly different between CRC and control groups. A risk model was constructed according to prognostic genes and verified by GSE39582. In addition, the nomogram exhibited superior predictive accuracy for CRC. Furthermore, immune analysis results indicated a notable positive correlation between risk score and the scores of immune (R = 0.486), stromal (R = 0.309), and ESTIMATE (R = 0.422). Immune checkpoint inhibitor ADORA2A (Cor = 0.483263) exhibited the strongest positive correlation with risk score. And MPP2 exhibited the most potent activating influence on the cell cycle pathway, whereas ANK3 demonstrated the most significant inhibitory effect within the apoptosis pathway.

CONCLUSIONS

A new JAK-STAT related CRC prognostic model was constructed and validated, which possessed an underlying predictive potential for CRC patients' prognosis and could potentially enhance tailored guidance for immunotherapy.

Causal association between type 1 diabetes and autoimmune cholestasis: A bi-directional Mendelian randomized study

Explore the causal relationship of risk between type 1 diabetes and primary biliary cholangitis (PBC) and primary sclerosing cholangitis (PSC). A causal association between type 1 diabetes and autoimmune liver disease remains ambiguous. This study explored potential causality between different autoimmune conditions, indicating that caution should be taken of the occurrence of autoimmune liver diseases in daily management of T1D patients. Genetic variants were extracted as instrumental variables from the genome-wide association study (GWAS) of PBC, PSC, type 1 diabetes (T1D), and type 2 diabetes (T2D). Associations between four primary liver enzymes, alkaline phosphatase (ALP), aspartate aminotransferase (AST), alanine transferase (ALT), and glutamyl transaminase (GGT), and blood glucose-related indicators such as 2h-glucose post-challenge (2hGlu), fasting glucose (FG), fasting insulin (FI), and glycated hemoglobin (HbA1c) were also evaluated (GWAS p-value < 5 × 10-8). A bi-directional two-sample Mendelian randomization (MR) design was used to assess causality between type 1 diabetes and autoimmune cholestasis. Genetic susceptibility to T1D increased the risk of PSC and PBC. Genetic susceptibility to T2D reduced the risk of PSC and showed no correlation with PBC. Genetically susceptibility to PBC increased the risk of T1D and showed no correlation with T2D. Genetically susceptibility to PSC did not impact the risk of T1D and T2D. T1D patients have an increased risk of PBC/PSC, but such causation is not mediated or explained by the altered blood glucose levels. A bi-directional causal association was identified between type 1 diabetes and autoimmune cholestasis. The findings provide new insight into the management of patients with these conditions.

Association between telomere length and idiopathic normal pressure hydrocephalus: a Mendelian randomization study

Objective

Idiopathic normal pressure hydrocephalus (iNPH) is highly prevalent among elderly individuals, and there is a strong correlation between telomere length and biological aging. However, there is limited evidence to elucidate the relationship between telomere length and iNPH. This study aimed to investigate the associations between telomere length and iNPH using the Mendelian randomization (MR) method.

Methods

The genetic variants of telomere length were obtained from 472,174 UK Biobank individuals. Summary level data of iNPH were acquired from 218,365 individuals of the FinnGen consortium. Five MR estimation methods, including inverse-variance weighting (IVW), MR-Egger regression, weighted median, weighted mode and simple mode, were used for causal inference. Comprehensive sensitivity analyses were conducted to test the robustness of the results. In addition, multivariable MR was further implemented to identify potential mechanisms in the causal pathway from telomere length to iNPH.

Results

Genetically determined longer telomere length was significantly associated with decreased risk of iNPH (OR = 0.44, 95% CI 0.24-0.80; p = 0.008). No evident heterogeneity (Cochran Q = 138.11, p = 0.386) and pleiotropy (MR Egger intercept = 0.01, p = 0.514) were observed in the sensitivity analysis. In addition, multivariable MR indicated that the observed association was attenuated after adjustment for several vascular risk factors, including essential hypertension (IVW OR = 0.55, 95% CI 0.30-1.03; p = 0.061), type 2 diabetes (IVW OR = 0.71, 95% CI 0.09-5.39; p = 0.740) and coronary artery disease (IVW OR = 0.58, 95% CI 0.31-1.07; p = 0.082).

Conclusion

Our MR study revealed a strong negative correlation of telomere length with iNPH. The causal relationship might be driven by several vascular risk factors.

Novel Insights into Causal Effects of Serum Lipids and Apolipoproteins on Cardiovascular Morpho-Functional Phenotypes

Previous observational studies have explored the association between serum lipids, apolipoproteins, and adverse ventricular/aortic structure and function. However, whether a causal link exists is uncertain. This study employed a two-sample Mendelian randomization (MR), colocalization, reverse, and multivariable MR (MVMR) approach to examine the causal associations among five serum lipids, two apolipoproteins, and 32 cardiac magnetic resonance (CMR) traits. Utilizing single-nucleotide polymorphisms (SNPs) linked to serum lipids and apolipoproteins as instrumental variables. CMR traits from seven independent genome-wide association studies served as preclinical endophenotypes, offering insights into aortic and cardiac structure/function. The primary analysis utilized a random-effects inverse variance method (IVW), followed by sensitivity and validation analyses. In the primary IVW MR analyses, genetically predicted low-density lipoprotein cholesterol (LDL-C) levels were positively correlated with increased descending aorta strain (DAo strain) (β = 0.098; P = 2.69E-07) and ascending aorta strain (AAo strain) (β = 0.079; P = 5.19E-05). Genetically predicted high-density lipoprotein cholesterol (HDL-C) levels were positively correlated with left ventricular radial peak diastolic strain rate (LV-PDSRll) (β = 0.176; P = 2.89E-05) and the left ventricular longitudinal peak diastolic strain rate (LV-PDSRrr) (β = 0.059; P = 2.44E-06), and negatively correlated with left ventricular regional wall thickness (LVRWT). While apolipoprotein B (ApoB) levels were positively correlated with AAo strain (β = 0.076; P = 1.16E-05), DAo strain (β = 0.065; P = 2.77E-05). A shared causal variant was identified to demonstrate the associations of ApoB with AAo strain and DAo strain using colocalization analysis. Sensitivity analyses confirmed the robustness of these associations. Targeting lipid and apolipoprotein levels through interventions may provide novel strategies for the primary prevention of CVDs.

Common pitfalls in drug target Mendelian randomization and how to avoid them

BACKGROUND

Drug target Mendelian randomization describes the use of genetic variants as instrumental variables for studying the effects of pharmacological agents. The paradigm can be used to inform on all aspects of drug development and has become increasingly popular over the last decade, particularly given the time- and cost-efficiency with which it can be performed even before commencing clinical studies.

MAIN BODY

In this review, we describe the recent emergence of drug target Mendelian randomization, its common pitfalls, how best to address them, as well as potential future directions. Throughout, we offer advice based on our experiences on how to approach these types of studies, which we hope will be useful for both practitioners and those translating the findings from such work.

CONCLUSIONS

Drug target Mendelian randomization is nuanced and requires a combination of biological, statistical, genetic, epidemiological, clinical, and pharmaceutical expertise to be utilized to its full potential. Unfortunately, these skillsets are relatively infrequently combined in any given study.

Drug-target Mendelian randomisation applied to metabolic dysfunction-associated steatotic liver disease: opportunities and challenges

Metabolic dysfunction-associated steatotic liver disease (MASLD) has emerged as the most prevalent cause of chronic liver disease worldwide affecting over one-third of the adult population. Despite the recent evolution of new nomenclature and diagnostic criteria for MASLD, progress in drug development for this condition remains limited. This review highlights the potential of drug-target Mendelian randomisation (MR), a study design that leverages human genetics and genomics, for the discovery, repositioning and safety assessment of drug targets in MASLD. We summarised key aspects of designing and appraising a drug-target MR study, discussing its inherent assumptions and considerations for instrument selection. Furthermore, we presented real-world examples from studies in MASLD which focused on opportunities and challenges in identifying novel drug targets, repositing existing drug targets, informing adjunctive treatments and addressing issues in paediatric MASLD.

Addressing the credibility crisis in Mendelian randomization

BACKGROUND

Genome-wide association studies have enabled Mendelian randomization analyses to be performed at an industrial scale. Two-sample summary data Mendelian randomization analyses can be performed using publicly available data by anyone who has access to the internet. While this has led to many insightful papers, it has also fuelled an explosion of poor-quality Mendelian randomization publications, which threatens to undermine the credibility of the whole approach.

FINDINGS

We detail five pitfalls in conducting a reliable Mendelian randomization investigation: (1) inappropriate research question, (2) inappropriate choice of variants as instruments, (3) insufficient interrogation of findings, (4) inappropriate interpretation of findings, and (5) lack of engagement with previous work. We have provided a brief checklist of key points to consider when performing a Mendelian randomization investigation; this does not replace previous guidance, but highlights critical analysis choices. Journal editors should be able to identify many low-quality submissions and reject papers without requiring peer review. Peer reviewers should focus initially on key indicators of validity; if a paper does not satisfy these, then the paper may be meaningless even if it is technically flawless.

CONCLUSIONS

Performing an informative Mendelian randomization investigation requires critical thought and collaboration between different specialties and fields of research.

Causal Links Between Systemic Disorders and Keratoconus in European Population

PURPOSE

To establish the presence of a causal linkage between prevalent systemic diseases and keratoconus (KC).

DESIGN

Mendelian randomization (MR) analysis.

METHODS

After an exhaustive screening process, genetic variants linked to various systemic diseases were identified as instrumental variables at the genome-wide significance level. Subsequently, MR analyses were conducted to elucidate their potential causal connection with KC (N = 26,742). The encompassed systemic ailments comprise diabetes, hay fever/allergic rhinitis/eczema, obstructive sleep apnea, thyroid dysfunction, aortic aneurysm, major depressive disorder, inflammatory bowel disease (including Crohn's disease and ulcerative colitis), and mitral valve prolapse. Our study adheres to the principles of Strengthening the Reporting of Observational Studies in Epidemiology Using MR guidelines.

RESULTS

Using inverse variance weighting as the primary MR analysis method, our findings revealed that hay fever/allergic rhinitis/eczema (odds ratio, 10.144; 95% CI, 2.441-42.149; P = .001) and ulcerative colitis (odds ratio, 1.147; 95% CI, 1.054-1.248; P = .002) were associated with an increased risk of KC within the largest population under scrutiny. Conversely, the prolonged hyperglycemic state did not exhibit a potentially protective effect in delaying the pathogenesis of KC, and no correlation was observed between the two (odds ratio, 0.320; 95% CI, 0.029-3.549; P = .353). Also, obstructive sleep apnea, thyroid function, aortic aneurysm, major depressive disorder, Crohn's disease, and mitral valve prolapse did not exhibit a causal association with KC (P > .05 for all comparisons).

CONCLUSIONS

This study indicates an increased risk of KC related to hay fever/allergic rhinitis/eczema and ulcerative colitis, with diabetes not providing a protective effect. These findings may potentially contribute some insights to inform clinical interventions.

COVID-19 and retinal layer thickness: A bidirectional Mendelian randomization study

BACKGROUND

Observational studies have reported that COVID-19 is associated with alterations in retinal layer thickness, including changes in the ganglion cell inner plexiform layer (GCIPL) and retinal nerve fiber layer (RNFL). However, the causal relationships remain unknown. Therefore, we assessed the direction and strength of the causal relationship between COVID-19 and GCIPL and RNFL thicknesses using a bidirectional two-sample Mendelian randomization (MR) design.

METHODS

Data were obtained from a large-scale COVID-19 Host Genetics Initiative (Nsample = 6,512,887), GCIPL dataset (Ncase = 31,434), and RNFL dataset (Ncase = 31,434). The inverse-variance weighted (IVW) method is the primary approach used to estimate causal effects. MR Egger, weighted median, weighted mode, MR Egger (bootstrap), and penalized weighted median methods were applied. Sensitivity analyses were implemented with RadialMR, MRPRESSO, MR-Egger regression, Cochran's Q statistic, leave-one-out analysis, and the funnel plot.

RESULTS

Forward MR analysis revealed that genetically identified COVID-19 susceptibility significantly increased the risk of GCIPL thickness (OR = 2.428, 95 % confidence interval [CI]:1.493-3.947, PIVW = 3.579 × 10-4) and RNFL thickness (OR = 1.735, 95 % CI:1.198-2.513, PIVW = 3.580 × 10-3) after Bonferroni correction. Reverse MR analysis did not indicate a significant causal association between GCIPL and RNFL thicknesses and COVID-19 phenotypes. No significant horizontal pleiotropy was found in the sensitivity analysis.

CONCLUSIONS

The host genetic liability to COVID-19 susceptibility was causally associated with increased GCIPL and RNFL thicknesses. Documenting this association increases our understanding of the pathophysiological mechanisms underlying COVID -19 susceptibility in retinopathy.

Migraine and gastroesophageal reflux disease: Disentangling the complex connection with depression as a mediator

OBJECTIVE

Gastroesophageal reflux disease (GERD) and migraine are public health concerns worldwide. No observational study has conclusively elucidated the causal relationship between these two conditions. We employed Mendelian randomization (MR) methods to explore the potential causal links between GERD and migraine.

METHODS

Genome-wide association studies were subjected to MR to infer the causality between GERD and migraine. Bidirectional two-sample MR was performed to establish causal relationships. Multivariable MR analysis was conducted to adjust potential confounding factors, and mediation MR analysis was utilized to assess the role of depression between GERD and migraine as a mediator. We primarily utilized the inverse variance weighted method (IVW) and sensitivity analysis methods, including MR-Egger, weighted median, and leave-one-out methods. We assessed heterogeneity and pleiotropy to ensure the reliability of the results.

RESULTS

Bidirectional two-sample MR revealed a positive causal effect of GERD on migraine (IVW: OR = 1.49, 95% CI: 1.34-1.66, p = 3.70E-13). Migraine did not increase the risk of GERD (IVW: OR = 1.07, 95% CI: 0.98-1.17, p = 0.1139). Multivariable MR indicated that the positive causal effect of GERD on migraine remained after adjustment for factors, such as smoking, alcohol consumption, obesity, type 2 diabetes, and depression. Mediation MR revealed that depression mediated 28.72% of GERD's effect on migraine. MR analysis was supported by all sensitivity analyses and was replicated and validated in another independent dataset on migraine.

CONCLUSION

Our findings elucidate the positive causal effect of GERD on migraine and underscores the mediating role of depression in increasing the risk of migraine due to GERD. Effective control of GERD, particularly interventions targeting depression, may aid in preventing the occurrence of migraine. Future research should delve deeper into the specific pathophysiological mechanisms through which GERD affects migraine risk, facilitating the development of more effective drug targets or disease management strategies.

Genetic evidence for the causal effects of C-reactive protein on self-reported habitual sleep duration

Inflammatory responses to acute stimuli are proposed to regulate sleep, but the relationship between chronic inflammation and habitual sleep duration is elusive. Here, we study this relation using genetically predicted level of chronic inflammation, indexed by CRP and IL6 signaling, and self-reported sleep duration. By Mendelian randomization analysis, we show that elevated CRP level within <10 mg/L has a homeostatic effect that facilitates maintaining 7-8 h sleep duration per day - making short-sleepers sleep longer (p = 2.42 × 10-2) and long-sleepers sleep shorter (1.87 × 10-7); but it is not associated with the overall sleep duration (p = 0.17). This homeostatic effect replicated in an independent CRP dataset. We observed causal effects of the soluble interleukin 6 receptor and gp130 on overall sleep duration (p = 1.62 × 10-8, p = 2.61 × 10-58, respectively), but these effects disappeared when CRP effects were accounted for in the model. Using polygenic score analysis, we found that the homeostatic effect of CRP on sleep duration stems primarily from the genetic variants within the CRP gene region: when genetic variants outside of this region were used to predict CRP levels, the opposite direction of effect was observed. In conclusion, we show that elevated CRP level may causally facilitate maintaining an optimal sleep duration that is beneficial to health, thus updating our current knowledge of immune regulation on sleep.