b) Двунаправленная менделичная оценка причинно-следственной связи между уровнями лактата и множественными аутоиммунными заболеваниями.

Introduction

Autoimmune diseases are a diverse group of conditions characterized by immune dysfunction that result in abnormal reactivity of B cells and T cells to normal components of the host. Autoimmune diseases can affect nearly every organ system and can be classified as either systemic or organ-specific 1 . The occurrence of autoimmune diseases is almost independent of demographic characteristics, and female sex was their only common risk factors. In UK, 63.9% of new autoimmune disease patients from 2000 to 2019 were female 2 . Risk factors for various other autoimmune diseases are not consistent, and childhood, social ladder, season, and co-occurrence of autoimmune diseases may contribute separately to the development of these conditions 2 . The clinical manifestations of autoimmune diseases vary widely, from life-threatening acute organ failure to more subtle laboratory abnormalities 3 . It is precisely because of the diversity of autoimmune diseases that complicated their diagnosis and treatment, making it a challenge for the entire medical enterprise.

Recent research has highlighted the role of lactate, a byproduct of aerobic glycolysis, in the pathophysiology of various diseases. Lactate production is elevated under stress conditions, including trauma, infection, myocardial infarction, and heart failure, as well as in tumor cells 4 – 6 . At present, lactate level is considered as an indicator for the diagnosis, prognosis and efficacy evaluation of a variety of diseases in clinical practice, including a number of autoimmune diseases. Studies have reported altered lactate levels in autoimmune conditions such as multiple sclerosis (MS), type 1 diabetes (T1D), rheumatoid arthritis (RA), and inflammatory bowel disease (IBD). Lactate levels (REST) were elevated and lactate levels (MAX) were decreased in MS patients, and the blood lactate levels are different at rest and during vigorous exercise compared to healthy individuals 7 . Urinary lactate levels in patients with diabetic nephropathy (DKD) are higher than those in controls of the same age, suggesting lactate as a key element in DKD pathogenesis 8 . In RA, in vitro studies have demonstrated that hypoxic and acidic conditions in the synovial membrane lead to increased lactate levels, which, in turn, influence fibroblast and macrophage activity, IL-6 secretion, and metabolic processes via specific lactate transporters 9 . Additionally, lactate levels were found to correlate with inflammation and disease progression in IBD models 10 . Other limited but similar evidence of a relationship between autoimmune diseases and lactate levels has been found in related studies of CD 11 , SLE 12 , and UC 13 .

While these studies suggest a potential link between lactate and autoimmune diseases, they are subject to biases inherent in observational, in vitro, or animal research. Mendelian randomization (MR) which uses genetic variation as an instrumental variable, may offer a more robust approach to assess the causal relationship between lactate levels and autoimmune diseases 14 . MR design can help minimize confounding factors due to the random distribution of genetic variants and provide stronger evidence for causality. However, the direction of causality between lactate levels and autoimmune diseases remains unclear, with conflicting findings in the existing literature. To address this gap, we conducted a bidirectional two-sample MR study using data from genome-wide association studies (GWAS) to explore the causal association between autoimmune diseases and lactate levels, incorporating a series of sensitivity analyses to strengthen our findings.

Therefore, this study aims to clarify the causal relationship between autoimmune diseases and lactate levels, potentially resolving existing discrepancies and providing novel insights into the underlying mechanisms of autoimmune pathogenesis.

Methods

Study design

This study utilized a bidirectional two-sample MR approach to investigate the causal relationship between autoimmune diseases and lactate levels. The core design and key assumptions of the present MR study were demonstrated in Fig.  1 . This research was reported according to the “STrengthening the Reporting of Observational studies in Epidemiology using Mendelian Randomization” (STROBE-MR) statement 15 .The methodological foundation of this investigation was constructed upon three core principles integral to MR analysis:

Design of Experiments flowchart.

Relevance assumption: Requires a strong connection between IVs and exposures.

Independence assumption: No existence of correlation between IVs and any variables that affected both exposure and outcome.

Exclusion restriction assumption: Other than the effect on exposure, IVs did not alter outcomes by any other causal pathways.

Based on publicly available GWAS data, no additional ethical approval was required for our MR analysis.

Data source

GWAS data related to both autoimmune diseases 16 and lactate levels 17 were obtained from the IEU Open GWAS datasets (Updated to December 06, 2023). There were 11,590,399 lactate levels related SNPs from 114,806 European populations included in this MR analysis. As for autoimmune diseases, we selected 7 conditions which including MS, T1D, RA, IBD, CD, SLE, and UC. Related GWAS data were all from European populations to avoid the potential bias due to racial differences. Ethical approval and consent to participate was not required, and details of exposure and outcome data were shown in Table S1 .

Instrumental variable selection

SNPs associated with the genome-wide significance of autoimmune diseases and lactate levels ( p  < 5 * 10 − 8 ) and minor allele frequency (MAF) > 0.01 were screened 18 . Based on the standard of R 2  < 0.001 and window size = 10,000kb 19 , linkage disequilibrium effect (LD) between SNPs was eliminated. When primary SNPs were unavailable in the outcome dataset, suitable SNPs with a high LD effect (R 2  > 0.8) were identified to replace the original IV as proxy SNPs. According to the formula F = R 2 * (N-2)/(1-R 2 ), this MR study calculated the F-value of each SNP in IV as the basis for assessing IV intensity. To minimize weak instrument bias, SNPs with F-statistics less than 10 were excluded during instrument selection.

MR analysis

In this MR study, four commonly used MR methods were included: random-effects IVW, weighted mode 20 , WME 21 , and MR-Egger. Considering the weighted average of the inverse variance of each SNP, the odds ratio (OR) and 95% confidence interval (CI) was calculated by IVW to interpret the results of MR 22 . In addition, MR-Egger, WME and weighted mode methods were used to test the robustness of the IVW analysis results 21 . This is due to the fact that MR-Egger can accurately estimate the causal effect under multi-effect bias, while WME method analyzes the causal relationship between exposure and results from another perspective, assuming the validity of half IVs.

Sensitivity analysis

Sensitivity analysis involved in this study was mainly used to detect potential pleiotropy in MR Studies. Analysis methods included Cochran’s Q test, MR-Egger regression, MR pleiotropy residual sum and outlier (MR-PRESSO) and leave-one-out analysis 23 , 24 . When P  > 0.05 of Cochran’s Q test was obtained, the heterogeneity among IVs was considered to be low, that is, the estimates among IVs were random and had no significant impact on the IVW results. MR-Egger regression method was selected to evaluate the influence of the pleiotropy of genetic variation on the estimation of association effects. When the intercept term of MR-Egger regression approaches zero or has no statistical significance, it indicates that there was no existence of pleiotropy. Additionally, the MR pleiotropy residual sum and outlier (MR-PRESSO) method was employed to identify and adjust for outliers in the SNPs. SNPs identified as outliers with a P-value less than 0.05 were excluded from the analysis, thus mitigating the effects of pleiotropy and improving the accuracy of the causal inferences 23 . Leave one-out analysis was used to test the robustness and consistency of the results. Finally, based on MR-IVW, MR-Egger, WEM and weighted mode, reverse causality detection was performed to detect potential reverse causality with lactate levels as IVs 25 .

Statistical analysis

The analysis was performed using R version 4.0.5 and the “Two-sample MR” package designed for MR studies. Graphical representations such as forest plots, scatter plots, and funnel plots were utilized to visually interpret the data. Steiger filtering was also conducted to ensure the directionality of the association between lactate levels and autoimmune diseases. To account for the inclusion of exposure variables, P-values were adjusted for multiple comparisons using the False Discovery Rate (FDR) correction method. Associations with adjusted P-values < 0.05 were considered statistically significant 26 .

Results

Selection of IVs

At genome-wide association significance ( P  < 5 * 10 − 8 ), our MR Analysis selected 510 autoimmune diseases related SNPs as IVs. Among them, there were 72, 90, 45, 117, 88, 62 and 36 IVs related to MS, RA, SLE, IBD, CD, UC and T1D, respectively. The mean and range of F-statistics value was 81.52 (29.81 to 1269.22) for MS, 107.49 (29.98 to 1487.90) for RA, 96.77 (29.96 to 460.93) for SLE, 70.35 (29.86 to 500.59) for IBD, 77.51 (30.14, 489.55) for CD, 70.36 (30.46 to 408.12) for UC, 181.97 (30.00 to 1406.97) for T1D. In all case of the MS, RA, and SLE related SNPs, there was 1 SNP each that did not match information in the summary data (rs9393975, rs41316148, rs28361029). Meanwhile, 2, 2, 2, 1, 2, 3, and 3 weak related IVs was eliminated from MS, RA, SLE, IBD, CD, UC and T1D related SNPs. After that, a strong correlation between IVs and exposures was protected (F-values > 10). The specific information of the selected SNPs is shown in Table S2 .

In the reverse direction, 11 SNPs associated with lactate levels (mean F = 64.00, [30.81, 174.32]) met the criteria and were screened as IVs in reverse MR analysis. When using MS, IBD, CD, UC and T1D as outcomes, there were 3, 2, 2, 2, and 2 SNPs did not match information in the summary data respectively. When RA was the outcome, the reverse MR Analysis removes 1 weak instrumental variable. Detailed information of the selected SNPs is shown in Table S3 .

Causal effects of autoimmune diseases on lactate levels

The results of this MR analyses examining the causal relationship between 7 genetically proxied autoimmune diseases and lactate levels were displayed. As a result, no significant genetic association between any autoimmune diseases and lactate levels was discovered by IVW (SLE (OR: 1.00; 95% CI [0.99, 1.00]; P FDR = 0.303); IBD (OR: 1.01; 95% CI [1.00, 1.02]; P FDR = 0.303); CD (OR: 1.00; 95% CI [0.99, 1.01]; P FDR = 0.825); UC (OR: 1.01; 95% CI [1.00, 1.02]; P FDR =0.366); T1D (OR: 1.00; 95% CI [0.99, 1.00]; P FDR = 0.484); RA (OR: 1.00; 95% CI [1.00, 1.01]; P FDR = 0.484); MS (OR: 1.00; 95% CI [0.99, 1.01]; P FDR = 0.825)). While IVW analysis revealed no statistically significant associations between autoimmune diseases and lactate levels after FDR correction, exploratory analyses using WME and weighted mode methods identified nominally significant associations between IBD or RA and lactate levels. Both WME (OR: 1.01; 95% CI [1.00, 1.02]; P  = 0.01) and weighted mode (OR: 1.01; 95% CI [1.00,1.02]; P  = 0.008) suggested a potential association between RA and lactate levels. These findings, though not surviving multiple testing correction (P FDR > 0.05), may, may suggest potential biological links warranting further investigation. Meanwhile, the weighted mode results nominally supported the association between IBD and lactate levels (OR: 1.01; 95% CI [1.00, 1.04]; P  = 0.042; P FDR = 0.303). Consistent with the above results, the MR regression slopes and individual causal estimates of each of the SNPs were illustrated in the scatter plot (Fig.  2 ) and forest plot (Figure S1 ). Detailed genetic association analysis data were shown in Table  1 .

( a ). Scatter plot of causal association of SLE (exposure) on lactate levels (outcome); ( b ). Scatter plot of causal association of IBD (exposure) on lactate levels (outcome); ( c ). Scatter plot of causal association of CD (exposure) on lactate levels (outcome); ( d ). Scatter plot of causal association of UD (exposure) on lactate levels (outcome); ( e ). Scatter plot of causal association of T1D (exposure) on lactate levels (outcome); ( f ). Scatter plot of causal association of RA (exposure) on lactate levels (outcome); ( g ). Scatter plot of causal association of MS (exposure) on lactate levels (outcome);.

Table 1 Bidirectional MR results of causal association between lactate levels and autoimmune disease. Exposure Outcome N .SNPs Methods OR (95% CI)
P
FDR adjusted P Systemic lupus erythematosus Lactate levels 41 IVW 1.00 (0.99–1.00) 0.081 0.303 MR Egger 1.00 (0.99–1.01) 0.438 0.808 Weighted median 1.00 (0.99–1.00) 0.399 0.454 Weighted mode 1.00 (0.99–1.01) 0.849 0.916 Inflammatory bowel disease Lactate levels 114 IVW 1.01 (1.00–1.02) 0.087 0.303 MR Egger 1.01 (0.99–1.03) 0.356 0.808 Weighted median 1.01 (1.00–1.02) 0.058 0.174 Weighted mode 1.02 (1.00–1.04) 0.042 0.21 Crohn’s disease Lactate levels 85 IVW 1.00 (0.99–1.01) 0.802 0.825 MR Egger 1.00 (0.98–1.02) 0.693 0.808 Weighted median 1.00 (0.99–1.01) 0.946 0.947 Weighted mode 1.00 (0.99–1.02) 0.911 0.916 Ulcerative colitis Lactate levels 58 IVW 1.01 (1.00–1.02) 0.157 0.366 MR Egger 1.01 (0.98–1.04) 0.602 0.808 Weighted median 1.01 (0.99–1.02) 0.272 0.403 Weighted mode 0.99 (0.95–1.02) 0.469 0.888 Type 1 diabetes Lactate levels 33 IVW 1.00 (0.99–1.00) 0.285 0.484 MR Egger 1.00 (1.00–1.01) 0.601 0.808 Weighted median 1.00 (0.99–1.00) 0.217 0.403 Weighted mode 1.00 (1.00–1.00) 0.891 0.916 Rheumatoid arthritis Lactate levels 86 IVW 1.00 (1.00–1.01) 0.345 0.484 MR Egger 1.01 (1.00–1.02) 0.022 0.154 Weighted median 1.01 (1.00–1.02) 0.010 0.071 Weighted mode 1.01 (1.00–1.02) 0.008 0.11 Multiple sclerosis Lactate levels 70 IVW 1.00 (0.99–1.01) 0.825 0.825 MR Egger 1.00 (0.99–1.01) 0.873 0.873 Weighted median 1.01 (1.00–1.02) 0.229 0.403 Weighted mode 1.00 (0.99–1.01) 0.438 0.888 Lactate levels Systemic lupus erythematosus 8 IVW 1.70 (0.73–3.92) 0.216 0.505 MR Egger 3.67 (0.26–51.62) 0.372 0.889 Weighted median 1.56 (0.64–3.79) 0.332 0.603 Weighted mode 1.46 (0.52–4.08) 0.493 0.689 Lactate levels Inflammatory bowel disease 7 IVW 1.11 (0.78–1.58) 0.552 0.644 MR Egger 1.27 (0.41–3.91) 0.695 0.889 Weighted median 1.04 (0.74–1.48) 0.806 0.803 Weighted mode 1.03 (0.71–1.51) 0.866 0.867 Lactate levels Crohn’s disease 7 IVW 1.40 (0.88–2.23) 0.156 0.505 MR Egger 1.04 (0.23–4.78) 0.958 0.958 Weighted median 1.32 (0.84–2.07) 0.235 0.603 Weighted mode 1.27 (0.78–2.06) 0.368 0.689 Lactate levels Ulcerative colitis 8 IVW 1.04 (0.62–1.75) 0.868 0.868 MR Egger 1.42 (0.20–9.87) 0.735 0.889 Weighted median 0.79 (0.45–1.38) 0.409 0.603 Weighted mode 0.73 (0.37–1.41) 0.376 0.689 Lactate levels Type 1 diabetes 8 IVW 0.85 (0.53–1.36) 0.491 0.644 MR Egger 1.25 (0.31–5.12) 0.762 0.889 Weighted median 0.91 (0.50–1.65) 0.757 0.803 Weighted mode 1.12 (0.55–2.26) 0.766 0.867 Lactate levels Rheumatoid arthritis 9 IVW 0.83 (0.65–1.06) 0.143 0.505 MR Egger 1.24 (0.63–2.43) 0.549 0.889 Weighted median 0.87 (0.63–1.21) 0.419 0.603 Weighted mode 0.84 (0.52–1.36) 0.499 0.689 Lactate levels Multiple sclerosis 7 IVW 0.84 (0.56–1.28) 0.420 0.644 MR Egger 0.60 (0.19–1.86) 0.418 0.889 Weighted median 0.79 (0.47–1.33) 0.376 0.603 Weighted mode 0.78 (0.43–1.43) 0.459 0.689 IVW, Inverse variance weighted method.

Bidirectional MR results of causal association between lactate levels and autoimmune disease.

IVW, Inverse variance weighted method.

Causal effects of lactate levels on autoimmune diseases

To assess any reverse causality, we used lactate levels as exposure and 7 autoimmune diseases as outcome. We found no statistically significant effect of lactate levels difference on any autoimmune diseases by IVW (SLE (OR: 1.70; 95% CI [0.73, 3.92]; P FDR = 0.505); IBD (OR: 1.11; 95% CI [0.78, 1.58]; P FDR = 0.644); CD (OR: 1.40; 95% CI [0.88–2.23]; P FDR = 0.505); UC (OR: 1.04; 95% CI [0.62–1.75]; P FDR = 0.868); T1D (OR: 0.85; 95% CI [0.53–1.36]; P FDR = 0.644); RA (OR: 0.83; 95% CI [0.65, 1.06]; P FDR = 0.505); MS (OR: 0.84; 95% CI [0.56, 1.28]; P FDR = 0.644)). The weighted model, WME, and MR-Egger gave similar causal estimates in magnitude, and the MR-Egger found no horizontal pleiotropy effect during the analysis. The visualization results of reverse MR analysis and detailed analysis results were shown in Fig.  3 , Figure S2 and Table  1 .

( a ). Scatter plot of causal association of lactate levels (exposure) on SLE (outcome); ( b ). Scatter plot of causal association of lactate levels (exposure) on IBD (outcome); ( c ). Scatter plot of causal association of lactate levels (exposure) on CD (outcome); ( d ). Scatter plot of causal association of lactate levels (exposure) on UD (outcome); ( e ). Scatter plot of causal association of lactate levels (exposure) on T1D (outcome); ( f ). Scatter plot of causal association of lactate levels (exposure) on RA (outcome); ( g ). Scatter plot of causal association of lactate levels (exposure) on MS (outcome).

Sensitivity analyses

Directional testing conducted by MR Steiger suggests that there may be a potential reverse causality in the analysis of SLE as outcomes and lactate levels as exposures. As for rest of the analysis, bidirectional MR analysis did not reveal any causal effect of autoimmune diseases on the lactate levels difference and Steiger filtering further ensured directionality (Table  2 ).

Table 2 MR Steiger test of directionality. Exposure Outcome SNP_r2.exposure SNP_r2.outcome Correct causal direction Steiger pval Systemic lupus erythematosus Lactate levels 0.303261822 0.001611019 TRUE < 0.001 Inflammatory bowel disease Lactate levels 0.137111343 0.002179226 TRUE < 0.001 Crohn’s disease Lactate levels 0.169153 0.001339276 TRUE < 0.001 Ulcerative colitis Lactate levels 0.094888357 0.000865334 TRUE < 0.001 Type 1 diabetes Lactate levels 0.316138109 0.00121741 TRUE < 0.001 Rheumatoid arthritis Lactate levels 0.14552832 0.00123998 TRUE < 0.001 Multiple sclerosis Lactate levels 0.046423839 0.000965938 TRUE < 0.001 Lactate levels Systemic lupus erythematosus 0.006131174 0.003989493 TRUE 0.086547831 Lactate levels Inflammatory bowel disease 0.005625952 0.001430344 TRUE < 0.001 Lactate levels Crohn’s disease 0.005625952 0.002127393 TRUE < 0.001 Lactate levels Ulcerative colitis 0.005625952 0.001029935 TRUE < 0.001 Lactate levels Type 1 diabetes 0.005625952 0.008066655 FALSE < 0.001 Lactate levels Rheumatoid arthritis 0.006131174 0.000617323 TRUE < 0.001 Lactate levels Multiple sclerosis 0.004840144 0.000302961 TRUE < 0.001

MR Steiger test of directionality.

The summary results of the sensitivity and pleiotropy analysis were shown in Table  3 . Except when using T1D as exposure, significant heterogeneity was observed in all of other causal estimates of individual genetic variation by Cochran’s Q test statistics (almost all P  > 0.05 except T1D as exposure, P  = 0.194). Since the IVW results of random effects were selected as the main MR Analysis method, the influence of heterogeneity could be ignored, and the funnel plot of symmetrical distribution also proved the robustness of the results. (Table  3 and Figure S3 - S4 ). However, the result of MR-Egger regression shows that the MR results by using RA as exposure and lactate levels as outcomes may be affected by horizontal pleiotropy (MR-Egger Intercept: – 0.0025, P  = 0.018) (Table  3 ). No significant intercepts indicating directional pleiotropy were observed for other primary forward MR analyses (all P  > 0.05).

Table 3 Sensitivity and Pleiotropy analysis of causal association between lactate levels and autoimmune disease. Exposure Outcome Heterogeneity Pleiotropy Q statistic (IVW) P value MR-Egger Intercept P value Systemic lupus erythematosus Lactate levels 178.50 < 0.001 0.003587159 0.332 Inflammatory bowel disease Lactate levels 245.64 < 0.001 0.000471426 0.764 Crohn’s disease Lactate levels 152.59 < 0.001 -0.00098073 0.586 Ulcerative colitis Lactate levels 93.96 0.002 -0.000698135 0.771 Type 1 diabetes Lactate levels 38.65 0.194 -0.002511242 0.049 Rheumatoid arthritis Lactate levels 140.46 < 0.001 -0.002477147 0.018 Multiple sclerosis Lactate levels 110.42 0.001 1.57207E-05 0.988 Lactate levels Systemic lupus erythematosus 52.21 < 0.001 -0.081631635 0.293 Lactate levels Inflammatory bowel disease 70.09 < 0.001 -0.006993046 0.871 Lactate levels Crohn’s disease 61.78 < 0.001 0.011022045 0.833 Lactate levels Ulcerative colitis 45.52 < 0.001 -0.058956279 0.244 Lactate levels Type 1 diabetes 133.43 < 0.001 -0.064031839 0.542 Lactate levels Rheumatoid arthritis 35.87 < 0.001 -0.017522133 0.545 Lactate levels Multiple sclerosis 35.06 < 0.001 -0.005474229 0.903

Sensitivity and Pleiotropy analysis of causal association between lactate levels and autoimmune disease.

Additionally, we applied the MR-PRESSO method, aimed at detecting and rectifying potential horizontal pleiotropy and outliers. Outliers were discovered in almost all analysis, except when IBD or RA as exposures and lactate levels as outcomes. After removing the outliers, MR-PRESSO results indicated that no genetic association was observed between any autoimmune diseases and lactate levels in any analysis directions ( P  > 0.05) (Table S4 ).

To further explore the potential impact of outlying genetic variants, we performed a leave-one-out analysis, systematically removing each instrumental variable and recalculating the causal estimate. However, this analysis did not demonstrate any significant changes in the overall causal estimates for any of the exposure (Figures S5 - S6 ), indicating that the null findings were not influenced by any individual influential instrumental variable.

Discussion

To the best of our knowledge, this is the first study to combine seven different types of autoimmune diseases and lactate levels data to explore casual association between autoimmune diseases and lactate level difference using bidirectional two-sample MR method. Seven autoimmune diseases including MS, RA, SLE, IBD, CD, UC and T1D. All of these autoimmune diseases have been proved to be associated with lactate level change by previous epidemiological, in vitro, and animal studies, which was opposite with our MR analysis result. According to the bidirectional MR results, no statistically robust genetic causal association was discovered between 7 autoimmune diseases and lactate level difference in any direction after correction for multiple testing. However, exploratory analyses using WME and weighted mode methods did reveal nominally significant associations between RA and lactate levels and between IBD and lactate levels in the forward MR direction. While these findings did not survive FDR correction and require cautious interpretation, particularly given the potential pleiotropy detected in the RA-lactate analysis (discussed below), they may hint at potential weak biological links warranting further investigation with larger sample sizes or more sensitive methodologies.

Current research evidence often focuses on MS and T1D populations when exploring links with lactate. Biological plausibility exists for such links: In Multiple Sclerosis (MS), chronic inflammation within the central nervous system (CNS) creates a unique microenvironment potentially altering local lactate metabolism 7 , 27 , and mitochondrial dysfunction, a hallmark of MS progression, can lead to lactate accumulation 28 , 29 . Observational studies have indeed reported higher serum lactate in progressive MS 30 . However, our bidirectional MR analysis found no robust genetic evidence supporting a causal relationship between systemic lactate levels and MS risk, or vice versa. This discrepancy might arise because MR assesses systemic circulating lactate, which may not adequately reflect localized CNS metabolic changes critical to MS pathology. Furthermore, insufficient statistical power in the available GWAS datasets or confounding in observational studies could also contribute to the differing results.

Similarly, in Type 1 Diabetes (T1D), altered glucose metabolism and potential mitochondrial dysfunction, particularly in complications like diabetic nephropathy (DKD), provide a rationale for investigating lactate 8 . Observational studies have linked factors like impaired hypoglycemic awareness or hormonal influences to lactate dynamics in T1D 31 – 33 . Despite these potential connections, our MR analysis did not detect a significant causal effect between genetic liability to T1D and circulating lactate levels, nor in the reverse direction. Possible explanations include limitations in statistical power, the nature of the lactate GWAS capturing baseline systemic levels rather than dynamic responses, or non-causal explanations for previously observed associations.

Regarding Rheumatoid Arthritis (RA), the inflamed synovium is recognized as a high-lactate microenvironment where lactate plays complex roles in modulating fibroblast and macrophage function 9 . Lactate levels in synovial fluid have even been explored as a diagnostic marker 34 . Despite this established local pathophysiological link and potential biomarker utility, our MR analysis did not find robust genetic evidence for a causal relationship between RA susceptibility and systemic circulating lactate levels, or vice versa (after FDR correction). While exploratory analyses with WME and weighted mode hinted at a nominal association, this finding was not statistically robust after multiple testing correction and, critically, was potentially confounded by significant horizontal pleiotropy detected via the MR-Egger intercept test ( P  = 0.018). This suggests that systemic lactate levels, as captured by germline genetic prediction, may not be a primary causal factor for RA onset, or that the potential causal link is weak, bidirectional in a way MR cannot fully resolve with current instruments, or significantly biased by pleiotropy. The discrepancy between local synovial and systemic lactate roles warrants further investigation. It is hoped that with the accumulation of GWAS data, we can see more precise proof results in the future.

For Inflammatory Bowel Disease (IBD), including Crohn’s disease (CD) and Ulcerative Colitis (UC), animal models suggest links between intestinal lactate levels, gut permeability, and local inflammation 13 . However, consistent with limited direct human evidence, our MR analysis identified no significant causal association between genetic liability to IBD, CD, or UC and systemic circulating lactate levels, nor in the reverse direction (after FDR correction). A nominal signal for the effect of IBD on lactate via weighted mode did not survive correction. The lack of robust causal findings might reflect differences between localized gut mucosal lactate metabolism and systemic levels, insufficient GWAS power, or secondary, non-causal changes in lactate during active intestinal inflammation.

In the case of SLE, no direct causal evidence linking it to lactate, and only one study has observed serum lactate levels greater than 2 mM on admission or 24 h before admission to the intensive care unit (ICU) as a major risk factor for mortality increase 12 . Congruently, our MR analysis found no significant causal association between SLE susceptibility and systemic lactate levels, or vice versa.

Overall, the predominantly null causal findings from our comprehensive MR analysis contrast with some observational or mechanistic hypotheses linking lactate to these autoimmune diseases. This underscores the importance of using genetically informed methods to probe causality. The discrepancies observed could stem from several factors inherent to MR studies and the complexity of both lactate metabolism and autoimmune pathophysiology. Lactate dynamics are influenced by numerous factors beyond baseline genetic predisposition, including exercise, medication, diet, and crucially, the dynamic nature of autoimmune diseases themselves, with fluctuations between active flares and remission 7 , 35 . Our MR approach, based on GWAS of general disease susceptibility, cannot capture these temporal variations or stage-specific effects.

Several limitations should be acknowledged when interpreting our findings. First, while MR minimizes traditional confounding, residual confounding or unaccounted pleiotropy cannot be entirely excluded, although sensitivity analyses were largely reassuring. However, the significant MR-Egger intercept detected for the association between RA and lactate levels specifically indicates potential directional pleiotropy, requiring significant caution when interpreting any link involving RA. Second, due to our strict threshold limits, many of the genetic burdens of autoimmune disease and lactate levels were excluded during the IVs selection phase, which may lead to a failure outcome. Third, the use of summary-level GWAS data precluded stratification by clinical heterogeneity, such as disease activity (flare vs. remission) or severity. Since lactate levels can fluctuate with disease state, averaging effects across diverse patient subgroups might mask true associations present only in specific clinical contexts. Finally, our analyses were restricted to individuals of European ancestry, limiting the generalizability of the findings to other populations. Future research in diverse cohorts and studies incorporating detailed clinical data or utilizing individual-level data are needed to address these limitations and further clarify the complex relationship between lactate metabolism and autoimmune diseases.

In summary, we performed a bidirectional two-sample MR analysis for evidence of a putative causal relationship between autoimmune diseases and lactate levels by applying several genetic information methods. No causal association between them was discovered, with limited evidence points to possible effects of RA and IBD on lactate levels. Further studies are needed to explore the roles of these autoimmune diseases in lactate levels.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Supplementary Material 3

Supplementary Material 4

Supplementary Material 5

Supplementary Material 6

Supplementary Material 7

Supplementary Material 8

Supplementary Material 9

Supplementary Material 10

Supplementary Material 11

Author contributions

Hongna Song, Qizhi Fu and Ganqin Du carried out the studies, participated in collecting data, and drafted the manuscript. Hua Fan and Xiaofei Shi performed the statistical analysis and participated in its design. Yongjie Bai and Canfei Zhang participated in acquisition, analysis, or interpretation of data and draft the manuscript. All authors read and approved the final manuscript.

Funding

This study was supported by Joint Fund of Henan Provincial Science and Technology R&D Project (No. 242103810034), Henan Academy of Innovations in Medical Science “Three Hundreds” Plan (S20240092), Henan Provincial Young and Middle-aged Health Science and Technology Innovation Leading Talent Training Program (No. LJRC2024019) and Heluo Young Talents Support Program (2024HLTJ08).

Data availability

All data generated or analyzed during this study are included in this article and supplementary information files.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

This article is a mendelian randomization study. The data for this study were obtained from publicly available databases and published literature data and does not require ethical approval and written informed consent.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Hua Fan, Hongna Song and Qizhi Fu contributed equally to this work.

Contributor Information

Hua Fan, Email: fanhua19860208@gmail.com.

Xiaofei Shi, Email: xiaofeis@haust.edu.cn.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-02507-9.

References

Associated Data

Supplementary Materials

Supplementary Material 1

Supplementary Material 2

Supplementary Material 3

Supplementary Material 4

Supplementary Material 5

Supplementary Material 6

Supplementary Material 7

Supplementary Material 8

Supplementary Material 9

Supplementary Material 10

Supplementary Material 11

Data Availability Statement

All data generated or analyzed during this study are included in this article and supplementary information files.