SGLT2 inhibition,circulating biomarkers,and Alzheimer's disease: A Mendelian randomization study

Abstract

Background

Sodium-glucose cotransporter 2 (SGLT2) inhibitors is a novel category of medications for diabetes, exhibiting neuroprotective potential. However, evidence regarding whether the use of SGLT2 inhibitors effectively reduces the risk of Alzheimer's disease (AD) remains unclear.

Objective

Our study employed Mendelian randomization (MR) analysis to investigate potential causal relationships between SGLT2 inhibition, metabolites, and AD.

Methods

In our research, we used a two-sample MR method to explore the link between SGLT2 inhibitor use and AD, addressing both its late-onset and early-onset forms. Furthermore, we executed a two-step MR analysis to explore how circulating metabolites, primarily endogenous in nature due to SGLT2 inhibition, mediate the relationship between SGLT2 inhibition and AD. The genetic instruments for SGLT2 inhibition were pinpointed through their association with SLC5A2 gene expression and the decreased glycated hemoglobin (HbA1c) levels.

Results

Genetic analysis indicated that SGLT2 inhibition, which effectively reduces HbA1c by enhancing renal glucose excretion and improving glycemic control, was associated with a lower likelihood of developing AD for every 1 SD decrease in HbA1c (OR = 0.48, [0.36, 0.63], p < 0.001). Our MR analysis revealed that SGLT2 inhibition significantly affected 27 of the 123 metabolites examined, adhering to a Bonferroni correction threshold (p < 4.06 × 10-4). Among these 27 significant metabolites, citrate was also associated with AD, showing a significant association (0.81 [0.79, 0.83], p < 0.001).

Conclusions

The study provides strong evidence linking SGLT2 inhibition with a lower AD risk, highlighting citrate's potential mediating role for subsequent clinical research.

Keywords

Alzheimer's disease circulating biomarkers Mendelian randomization sodium-glucose cotransporter 2 inhibition

Introduction

Sodium-glucose cotransporter 2 (SGLT2) inhibitors, represent a notable advancement in managing type 2 diabetes mellitus (T2DM), functioning by inhibiting SGLT2 in the kidneys, thus reducing glucose reabsorption and boosting urinary glucose excretion.¹ Their utility extends beyond glucose regulation, showing effectiveness in improving cardiovascular and renal outcomes.^2,3 These drugs have a notable influence on various circulating metabolites, although the precise nature of these interactions remains somewhat elusive, with research yielding mixed results.^4,5

Given the broader implications of such systemic effects, researchers have begun exploring the potential of repurposing medications like SGLT2 inhibitors for neurodegenerative diseases, including Alzheimer's disease (AD). AD is a disease marked by amyloid-β plaques accumulation and tau protein tangles, resulting in cognitive decline and memory loss.⁶ It manifests in two variations: early-onset AD (EOAD), linked to genetic factors, and late-onset AD (LOAD), which occurs more sporadically in the elderly population. Despite similar pathological markers, they exhibit distinct disease progressions.^7,8

Recent research has pointed to a potential connection between T2DM and a heightened risk of AD, possibly mediated by insulin resistance and disrupted glucose metabolic process in the brain.⁹ Given their anti-inflammatory and vasculoprotective qualities, SGLT2 inhibitors might offer a therapeutic avenue in AD management.¹⁰ Establishing a direct causal link, however, remains a challenge, largely due to the complexity of reverse causality.

Mendelian randomization (MR) offers a robust approach to explore potential cause-and-effect links between exposures, like SGLT2 inhibition, and outcomes such as AD. By leveraging genetic variants as instrumental variables (IVs), MR minimizes reverse causation and confounding, simulating the effects of randomized controlled trials.¹¹ This approach is particularly relevant in disentangling the complex interplay between metabolic disorders and neurodegenerative diseases and could be instrumental in identifying more definitive causal links. This is crucial for developing targeted therapeutic strategies for AD.^12,13

In our study, employing MR, we first rigorously examine the relationship between SGLT2 inhibition and AD. This initial investigation sets the foundation for a deeper exploration into how circulating metabolites may mediate this relationship. This offers a fresh perspective on the future therapeutic impact of SGLT2 inhibitors on AD, potentially paving the way for novel treatment strategies in this challenging and increasingly prevalent condition.

Methods

Study design

In our study (Figure 1): 1) We first identify genetic variants that act as indicators for the inhibitory effects of SGLT2; 2) We then select 123 metabolites as potential mediators from GWAS summary statistics; 3) Our analysis includes three outcomes: AD, LOAD, and EOAD; 4) We employ a two-step MR approach to determine the causal connections. Initially, we assess the effect of SGLT2 inhibition on the selected outcomes. Subsequently, we assess the influence of SGLT2 inhibition on the metabolites. Finally, we evaluate the significant metabolites for their role in AD.

Our commitment to the Strengthening the Reporting of Observational Studies in Epidemiology using Mendelian Randomization (STROBE-MR) guidelines guarantees the clarity and rigorous documentation of our research's methods and findings.¹⁴

Genetic instruments for SGLT2 inhibition

In our study, we identified genetic variants associated with SGLT2 inhibition through a detailed four-step process. We began by using Genotype-Tissue Expression (GTEx)¹⁵ and eQTLGen Consortium¹⁶ data to identify variants linked to the mRNA levels of the SLC5A2 gene, which encodes SGLT2. We then examined how these variants influenced glycated hemoglobin (HbA1c) levels, as a marker for glucose reduction, selecting variants with significant associations. Only variants with a significant correlation with HbA1c (p < 1 × 10⁻⁴) were advanced for further analysis.

The analysis of HbA1c leveraged genome-wide association study (GWAS) data from 344,182 individuals without diabetes of European ancestry, sourced from the UK Biobank (Supplemental Table 1). Our analysis included a colocalization assessment to ensure shared causality between SLC5A2 expression and HbA1c levels.¹⁷ Variants with a high posterior probability (>70%) for shared causality were considered supportive of colocalization. Finally, we grouped these variants based on linkage disequilibrium (LD) (r2 < 0.8 within a 250 kb range), referencing the 1000 Genomes European panel.

Genetically determined metabolites

For metabolites with genetic determinants, we sourced GWAS summary data for 123 circulating metabolites, drawn from an analysis across 14 cohorts.^18,19 This dataset was initially compiled and analyzed through a comprehensive GWAS to determine the levels of various metabolites in human serum.¹⁸ The original research considered adjustments for the duration of fasting. These metabolites cover a wide array of systemic metabolic processes, categorized into 12 classes, including a range of compounds from fatty acyls to steroids. We accessed data for these biomarkers from the IEU Open-GWAS Project database, under the identifier ‘met-c’ (Supplemental Table 1), concentrating on genetic variants that achieved genome-wide significance (p < 5 × 10⁻⁸) and were not in linkage disequilibrium (LD r² < 0.001 over a span of 10,000 kb).

Outcomes

For our primary outcome, we employed GWAS summary statistics from a significant 2019 meta-analysis by a major consortium, involving 21,982 cases of AD and 41,944 controls. This meta-analysis, comprising groups like Alzheimer Disease Genetics Consortium (ADGC), the European Alzheimer's Disease Initiative (EADI), the Cohorts for Heart and Aging Research in Genomic Epidemiology Consortium (CHARGE), and the Genetic and Environmental Risk in AD/Defining Genetic, Polygenic, and Environmental Risk for Alzheimer's Disease Consortium (GERAD/PERADES), identified risk loci and pathways related to key factors in AD.²⁰ Subsequently, different forms of AD, including LOAD and EOAD, were obtained from the tenth round of results (R10) in the FinnGen database.²¹ Specifically, for late-onset AD, the dataset comprised 7308 cases and 183,753 controls, while for early-onset AD, there were 1451 cases matched with an equal number of controls.

Statistical analyses

We conducted a two-sample MR analysis to assess the impacts of SGLT2 inhibition on AD. We applied the Mendelian Randomization Pleiotropy RESidual Sum and Outlier (MR-PRESSO)²² to identify and adjust for potential horizontal pleiotropy and heterogeneity, excluding outlier genetic variants prior to the MR analysis. The primary method we used was the Inverse Variance Weighted (IVW) approach, ideal when all the genetic variants are valid instruments.²³

In the mediation MR analysis, we adopted a sequential approach to analyze how circulating metabolites mediate the connection between SGLT2 inhibition and AD. Initially, we calculated the total effect of genetic variations on all outcomes (β0), without distinguishing between direct and mediating factor-mediated impacts. Subsequently, we assessed the influence of genetic variations on circulating metabolites (β1). Finally, we assessed the effect of significant mediating metabolites on AD (β2). The mediation effect was computed by multiplying β1 and β2.²⁴ The Delta method was used to calculate the 95% confidence intervals for these mediation proportions.²⁵

During our sensitivity analysis within the MR investigation of the effect of SGLT2 inhibition on the outcomes, we utilized multiple techniques such as MR-Egger, weighted median, simple mode, and weighted mode methods to ensure the reliability of our results. In our sensitivity analysis, the MR-Egger method played a crucial role in detecting directional pleiotropy by inspecting its intercept.²⁶ A non-zero intercept in MR-Egger would indicate such pleiotropy, potentially affecting the IVW estimates. Meanwhile, the weighted median approach was effective when a majority (at least 50%) of the genetic instruments were valid.²⁷ Moreover, the simple mode and weighted mode approaches proved valuable when the predominant cluster of instruments showed no signs of horizontal pleiotropy.^28,29

We performed heterogeneity assessments using the MR Egger and IVW methods to examine the variability among the genetic instruments; a p-value over 0.05 suggested no significant heterogeneity. For horizontal pleiotropy assessment, MR Egger regression was employed, with a p-value greater than 0.05 denoting its absence.

For selecting IVs, we determined each IV's variance explained and calculated the F-statistic using the equation F = (n−k−1 / k) * (R² / 1−R²), where n represents the sample size, k denotes the number of genetic instruments, and R² indicates the proportion of variance in the exposure explained by these instruments.³⁰ An instrument was considered strong if its F-statistic exceeded 10. To verify the robustness of our findings, we employed the leave-one-out approach, removing each SNP in turn to detect any substantial impacts on the results.³¹

We used the TwoSampleMR and MRPRESSO packages in the R software (version 4.2.2) to investigate MR analysis, considering a two-sided p-value for statistical significance and applying the Bonferroni correction for multiple comparisons. Nevertheless, for estimating the effects of SGLT2 inhibition on AD, a p-value under 0.05 was regarded as significant.

Standard protocol approvals, registrations, and patient consents

This study is based on GWAS summary statistics from meta-analysis. All individual participants in the constituent cohorts of the meta-analysis had given written informed consent, and the protocols for each cohort received approval from the respective local institutional review board.

Results

Influence of SGLT2 inhibition on AD

In our research, we utilized 10 independent SNPs as genetic markers for SGLT2 inhibition, all demonstrating an F statistic above 10 (Supplemental Table 1). The MR assessments indicated a notable link between SGLT2 inhibition and a decreased incidence of AD (Figure 2). For AD, the odds ratio (OR) for each standard deviation decrease in HbA1c was 0.48 (95% CI: 0.36 to 0.63) (Figure 2). These results were corroborated through sensitivity analyses using MR-PRESSO, which showed no signs of heterogeneity (Q = 1.312, p = 0.998) or horizontal pleiotropy (Egger intercept = −0.0300, p = 0.374) (Table 1). Consistent outcomes were observed for both LOAD and EOAD. For LOAD, the OR was 0.29 (95% CI: 0.14 to 0.60, p = 0.001), and for EOAD, the OR was also 0.29 (95% CI: 0.09 to 0.89, p = 0.03) (Table 1). There was no significant heterogeneity for LOAD (Q = 4.94, p = 0.839) or EOAD (Q = 3.034, p = 0.963), nor was there evidence of horizontal pleiotropy for LOAD (Egger intercept = -0.016, p = 0.719) or EOAD (Egger intercept = 0.077, p = 0.391). The leave-one-out approach also verified the consistency of these outcomes for AD (Figure 3).

Figure 1.

Study Design. (A) Two-sample Mendelian randomization (MR) analyses were conducted to assess the impact of SGLT2 inhibition on AD-related outcomes. Genetic variants representing the effect of SGLT2 inhibition were selected, and AD-related outcomes, including AD, LOAD, and EOAD, were chosen as outcomes. Summary data for exposure and outcomes were obtained from relevant meta-analyses of genome-wide association studies. The primary analysis employed the generalized inverse variance-weighted approach, with several sensitivity analyses performed. (B) To establish a causal relationship, three conditions were required: (1) instrumental variables should not be associated with confounders (dashed line), (2) instrumental variables should be linked to the exposure (solid line), and (3) instrumental variables should not directly affect the outcome (dashed line). (C) The two-step MR method was employed to estimate the mediation effect. SGLT2: Sodium-glucose cotransporter 2; HbA1c: glycated hemoglobin level; AD: Alzheimer's disease; LOAD: late-onset Alzheimer's disease; EOAD: early-onset Alzheimer's disease.

Figure 2.

Mr estimates of the association between genetically proxied SGLT2 inhibition and Alzheimer's disease. Estimates represent odds ratios (95% CIs) for Alzheimer's disease per SD increase in genetically predicted SGLT2 inhibition. OR: Odds ratio; AD: Alzheimer's disease; LOAD: Late-onset Alzheimer's disease; EOAD: Early-onset Alzheimer's disease.

Figure 3.

Leave-one-out sensitivity analysis of SGLT2 inhibition on AD, LOAD, EOAD. (A) AD. (B) LOAD. (C) EOAD. The leave-one-out method is used to evaluate the excessive effect of a single SNP on MR analysis if the comprehensive effect of the remaining SNPs is consistent with the main effect after removing one SNP. SNP: single nucleotide polymorphisms; SGLT2: sodium-dependent glucose transporter 2; AD: Alzheimer's disease; LOAD: Late-onset Alzheimer's disease; EOAD: Early-onset Alzheimer's disease.

Table 1.

Mr analysis of SGLT2 inhibition's impact on Alzheimer's disease (AD) related outcomes.

Outcome	Method	OR (95%CI)	p	Q statistic	p-heterogeneity	Egger intercept	p-intercept
AD	IVW	0.48 (0.36–0.63)	1.59E-07	1.312	0.998
	MR Egger	4.28 (0.04–439.67)	0.555	0.426	1.000	–0.0300	0.374
	Weighted median	0.55 (0.23–1.34)	0.192
	Simple mode	0.54 (0.15–1.98)	0.375
	Weighted mode	0.59 (0.18–1.94)	0.406
LOAD	IVW	0.29 (0.14–0.60)	0.001	4.940	0.839
	MR Egger	0.94 (0.00–460.28)	0.984	4.801	0.779	–0.016	0.719
	Weighted median	0.44 (0.12–1.68)	0.232
	Simple mode	0.62 (0.07–5.93)	0.692
	Weighted mode	0.59 (0.10–3.51)	0.574
EOAD	IVW	0.29 (0.09–0.89)	0.030	3.034	0.963
	MR Egger	0.00 (0.00–223.60)	0.306	2.212	0.974	0.077	0.391
	Weighted median	0.26 (0.02–3.07)	0.284
	Simple mode	0.30 (0.01–13.19)	0.547
	Weighted mode	0.25 (0.01–9.95)	0.483

This table presents odds ratios (OR), 95% confidence intervals (CI), and p-values derived from various MR analysis methods to assess the impact of SGLT2 inhibition on AD. Heterogeneity for the inverse-variance weighted (IVW) methods was assessed using Cochran's Q statistic, while the MR-PRESSO method utilized the global test. Significance was determined at a threshold of p < 0.05.

AD: Alzheimer's disease; LOAD: Late-onset Alzheimer's disease; EOAD: Early-onset Alzheimer's disease; IVW: Inverse-variance weighted; p-heterogeneity: p-value for the heterogeneity test; p-intercept: p-value for the MR-Egger regression intercept.

Mediation effect

Mediation MR of analysis of SGLT2 inhibition, metabolite levels, and AD

Our mediation MR assessment scrutinized the influence of SGLT2 inhibition on metabolite concentrations in the bloodstream and their subsequent effect on AD (Figure 4A). Out of 123 metabolites analyzed, 27 showed significant associations after applying the stringent Bonferroni correction (p value threshold = 4.06 × 10e⁻⁴ [0.05/123]). These metabolites encompass diverse classes, highlighting the broad metabolic modulation by SGLT2 inhibitors. Notably, several lipid-related metabolites displayed negative associations, marked by beta values less than zero. This was particularly evident in subclasses of VLDL and HDL particles. Certain amino acids, such as Alanine, Isoleucine, and Tyrosine, also showed decreased levels. Conversely, a positive correlation was observed for metabolites like Lactate and amino acids like Glycine, suggesting their upregulation upon SGLT2 inhibition.

Figure 4.

The forest plot of showing the effects of SGLT2 inhibition on circulating metabolites and the effects of metabolites on Alzheimer's disease. (A) The effects of SGLT2 inhibition on 27 out of 123 metabolites, which showed significant association with SGLT2 inhibition (Bonferroni-corrected P value threshold = 4.06 × 10−4 [0.05/123]). (B) The effects of these 27 metabolites on Alzheimer's disease (Bonferroni corrected p value threshold = 0.0019 [0.05/27]). CI: confidence interval; OR: odds ratio.

Further, we assessed the influence of these 27 metabolites on AD risk (Figure 4B). Notably, Citrate emerged as the sole metabolite with a robust inverse correlation with AD risk (OR = 0.81 [0.79, 0.83], p < 0.001), surpassing the Bonferroni-adjusted significance threshold of 0.0019 ([0.05/27]). Citrate accounted for a mediated portion of 20.15%, with the 95% confidence interval spanning 9.84% to 30.72%, in the total effect size noted. This result was further corroborated by MR-PRESSO analysis, which detected no signs of heterogeneity or horizontal pleiotropy. Notably, the presence of ‘NA’ in rows indicates insufficient SNP data to conduct MR analyses for those specific metabolites. Additionally, Lactate was not further analyzed due to an insufficient number of SNPs, thereby not fulfilling the criteria for the sensitivity assessment. The genetic variants for the 27 metabolites were determined to be robust (All F statistics > 10) (Supplemental Table 3).

Discussion

Key outcomes

Our research explored the links between genetic proxies for SGLT2 inhibition and the risk of AD. Additionally, we examined how circulating metabolites might mediate the relationship between SGLT2 inhibition and AD. These findings suggest that genetic alterations affecting SGLT2 inhibition are correlated with decreased risks AD. Moreover, our findings suggest the potential mediating role of citrate in the effect of SGLT2 inhibition on AD.

The association between SGLT2 inhibition and AD

Given the overlapping pathological elements between T2DM and AD, and the broad expression of SGLTs in brain tissue, there is a compelling argument for exploring the use of SGLT2 inhibitors in AD therapy.³² The putative neuroprotective properties of SGLT2 inhibitors have gathered attention for their potential to lower the likelihood of dementia and bolster cognitive capabilities via mechanisms such as reducing blood pressure, enhancing cardiac health, modulating inflammation, promoting weight reduction, and slowing the progression of atherosclerotic plaques.^33,34 Preclinical research in animal models demonstrates that SGLT2 inhibitors, specifically empagliflozin, can decrease Aβ plaque accumulation, limit brain cortex atrophy, lessen neuronal loss, and improve cognitive functions in AD model mice.³⁵ In humans, evidence is emerging, with studies showing that older individuals with T2DM taking SGLT2 inhibitors exhibit a lower dementia risk compared to those on other anti-diabetic drugs.³⁶ Additionally, a prospective cohort study suggested that prolonged treatment with SGLT2 inhibitors may lead to cognitive enhancements in overall cognitive assessment scores, as well as in specific areas like language and executive functioning among individuals with T2DM.³⁷

In our study, the findings that SGLT2 inhibition exerts a negative causal effect on both late-onset and early-onset AD. The observed uniform negative causal effect of SGLT2 inhibition on both AD types indicates a possible universal influence of disrupted glucose metabolism in the disease's pathology, regardless of onset age. This aligns with emerging evidence highlighting glucose metabolism abnormalities in AD brains, even in the absence of diabetes.³⁸ Furthermore, these findings underscore the potential of modulating systemic and cerebral glucose metabolism as a therapeutic strategy for AD, highlighting the need to consider metabolic interventions in future AD research and treatment paradigms.³⁹ Moreover, the consistent effect across AD subtypes emphasizes the importance of acknowledging disease heterogeneity in AD therapeutic exploration. It suggests that interventions targeting metabolic pathways could have broad applicability in AD, offering new avenues for research and therapeutic development.

The relationship between SGLT2 inhibition and metabolite levels

Previous research has provided mixed views on the impact of SGLT2 inhibitors on blood metabolites.^40,41 We observed alterations span a range of metabolic classes, reflecting the diverse biological roles of SGLT2 inhibitors. The negative beta values for VLDL and HDL subclasses, along with specific amino acids, suggest a lipid-lowering effect, consistent with findings from previous studies on the lipid-modifying effects of SGLT2 inhibitors.³ These changes in lipid profiles could potentially confer cardiovascular benefits, as reduced levels of certain VLDL and HDL particles have been associated with decreased atherosclerotic risk.⁴²

The reduction in amino acids such as Alanine, Isoleucine, and Tyrosine upon SGLT2 inhibition could be indicative of altered protein metabolism or energy utilization.⁴³ These changes might relate to improved insulin sensitivity and glucose utilization, as amino acid metabolism has been linked to insulin resistance and T2DM pathogenesis.⁴⁴ The decreased citrate levels observed may suggest a potential shift in mitochondrial energy metabolism, which could influence ATP production that is crucial for neuronal function. The elevation of glycine levels, observed as a positive beta value, may reflect changes in gluconeogenesis and nitrogen balance, which can be affected by SGLT2 inhibition due to its effects on renal glucose production and amino acid absorption.⁴⁵

The mediation role of blood metabolites in the link between SGLT2 inhibition and AD

The influence of circulating metabolites as mediators presents a complex network of interactions that could modulate the risk of AD following SGLT2 inhibition. Previous research has demonstrated both protective and detrimental effects of various metabolites on AD risk, indicating a multifaceted relationship between metabolism and neurodegeneration.^46,47 Our findings suggest that SGLT2 inhibitors might play a complex role in AD risk modulation. Overall, the inhibition may directly reduce AD risk by influencing glucose metabolism and insulin sensitivity, both of which are associated with the development of AD.⁹ However, this protective effect appears to be moderated by interactions with metabolic products such as citrate. Specifically, while increased citrate levels are generally associated with a decreased risk of AD,⁴⁷ SGLT2 inhibition unfortunately leads to a reduction in citrate levels, thereby potentially mitigating its beneficial effects on AD risk. Therefore, future pharmaceutical developments could consider combining SGLT2 inhibitors with treatments aimed at elevating citrate levels to enhance protective effects against AD. Additionally, although lactate was not analyzed further in our results due to insufficient SNPs, it is important to recognize lactate as a critical bioenergetic fuel for microglia and as a modulator of their proinflammatory activation.⁴⁸ Further research into lactate's role in AD is warranted to fully understand its therapeutic potential. Longitudinal studies and randomized controlled trials could elucidate whether the modulation of these metabolites by SGLT2 inhibitors has long-term benefits or drawbacks for cognitive health. Moreover, investigating how SGLT2 inhibition affects metabolite concentrations and AD biomarkers could shed light on the viability of SGLT2 inhibitors as treatments for AD.

Limitation

Our research utilizes a strong two-sample MR method to explore the link between SGLT2 inhibition and AD in the general population. This methodology, along with an extensive examination of 123 circulating metabolites and various genetic instruments, enhances the reliability of our insights into the impacts of SGLT2 inhibition and the mediating function of these metabolites. However, several limitations are noteworthy.

First, the genetic information, predominantly sourced from European-descended individuals, might restrict the generalizability of our results to various ethnic groups. Then, while the study identifies associations, it does not fully explain the complex interactions between circulating metabolites and their effect on AD. Also, a potential overlap in the Finnish subpopulations used for circulating metabolites and outcomes might introduce a degree of sample overlap bias, potentially skewing the causality analysis. Additionally, our study considers a limited selection of metabolites, which does not encompass all metabolic pathways relevant to our hypothesis, a limitation primarily due to the lack of comprehensive GWAS data for all metabolites. Finally, current study does not explore how genetic variability and environmental factors may jointly influence AD risk in the context of SGLT2 inhibition. Future analyses should incorporate more robust methods and sensitivity tests to better verify causal relationship. These aspects underscore the necessity for more nuanced research to comprehensively understand the intricate link between SGLT2 inhibition, circulating metabolites, and AD risk.

Conclusions

This research supported a relationship between genetic indicators of SGLT2 inhibition and AD, highlighting citrate's mediating role in this connection. The findings furnish genetic proof of how SGLT2 inhibition may decrease the risk of AD, offering valuable information for subsequent mechanistic and clinical research.

Supplemental Material

sj-docx-1-alz-10.1177_13872877241309674 - Supplemental material for SGLT2 inhibition, circulating biomarkers, and Alzheimer's disease: A Mendelian randomization study

Supplemental material, sj-docx-1-alz-10.1177_13872877241309674 for SGLT2 inhibition, circulating biomarkers, and Alzheimer's disease: A Mendelian randomization study by Hao Yang, Yuye Ning, Meilin Chen and Jianping Jia in Journal of Alzheimer's Disease

Footnotes

Acknowledgments

We would like to thank all the individuals who contributed to the GWASs that formed the basis of this research, as well as to the researchers who have facilitated access to this data through public repositories.

ORCID iD

Jianping Jia

Author contributions

Hao Yang (Conceptualization; Data curation; Methodology; Writing – original draft; Writing – review & editing); Yuye Ning (Writing – original draft; Writing – review & editing); Meilin Chen (Writing – original draft; Writing – review & editing); Jianping Jia (Writing – review & editing).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the Key Project of the National Natural Science Foundation of China (U20A20354); Beijing Brain Initiative from Beijing Municipal Science & Technology Commission (Z201100005520016, Z201100005520017); STI2030-Major Projects (No.2021ZD0201802); the National Key Scientific Instrument and Equipment Development Project (31627803); the Key Project of the National Natural Science Foundation of China (81530036).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

The data utilized in this research are confined to the GWAS summary statistics mentioned previously, accessible from the authors of the cited studies or from publicly available sources as specified.

Supplemental material

Supplemental material for this article is available online.

References

Scheen

. Pharmacodynamics, efficacy and safety of sodium-glucose co-transporter type 2 (SGLT2) inhibitors for the treatment of type 2 diabetes mellitus. Drugs 2015; 75: 33–59.

Packer

Anker

Butler

, et al. Cardiovascular and renal outcomes with empagliflozin in heart failure. N Engl J Med 2020; 383: 1413–1424.

Zinman

Wanner

Lachin

, et al. Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes. N Engl J Med 2015; 373: 2117–2128.

Liu

Sridhar

Montemayor

, et al. Changes in plasma and urine metabolites associated with empagliflozin in patients with type 1 diabetes. Diabetes Obes Metab 2021; 23: 2466–2475.

Katsuyama

Hamasaki

Adachi

, et al. Effects of sodium-glucose cotransporter 2 inhibitors on metabolic parameters in patients with type 2 diabetes: a chart-based analysis. J Clin Med Res 2016; 8: 237–243.

Graff-Radford

Yong

KXX

Apostolova

, et al. New insights into atypical Alzheimer's disease in the era of biomarkers. Lancet Neurol 2021; 20: 222–234.

Bateman

Xiong

Benzinger

, et al. Clinical and biomarker changes in dominantly inherited Alzheimer's disease. N Engl J Med 2012; 367: 795–804.

Jack Jr

Bennett

Blennow

, et al. NIA-AA Research framework: toward a biological definition of Alzheimer's disease. Alzheimers Dement 2018; 14: 535–562.

Biessels

Reagan

. Hippocampal insulin resistance and cognitive dysfunction. Nat Rev Neurosci 2015; 16: 660–671.

10.

Rayego-Mateos

Rodrigues-Diez

Fernandez-Fernandez

, et al. Targeting inflammation to treat diabetic kidney disease: the road to 2030. Kidney Int 2023; 103: 282–296.

11.

Emdin

Khera

Kathiresan

. Mendelian randomization. JAMA 2017; 318: 1925–1926.

12.

Dineley

Jahrling

Denner

. Insulin resistance in Alzheimer's disease. Neurobiol Dis 2014; 72: 92–103.

13.

Davey Smith

Hemani

. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet 2014; 23: R89–R98.

14.

Skrivankova

Richmond

Woolf

BAR

, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: the STROBE-MR statement. JAMA 2021; 326: 1614–1621.

15.

GTEx Consortium. The GTEx consortium atlas of genetic regulatory effects across human tissues. Science 2020; 369: 1318–1330.

16.

Võsa

Claringbould

Westra

, et al. Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat Genet 2021; 53: 1300–1310.

17.

Zuber

Grinberg

Gill

, et al. Combining evidence from Mendelian randomization and colocalization: review and comparison of approaches. Am J Hum Genet 2022; 109: 767–782.

18.

Kettunen

Demirkan

Würtz

, et al. Genome-wide study for circulating metabolites identifies 62 loci and reveals novel systemic effects of LPA. Nat Commun 2016; 7: 11122.

19.

Hagenbeek

Pool

van Dongen

, et al. Heritability estimates for 361 blood metabolites across 40 genome-wide association studies. Nat Commun 2020; 11: 39.

20.

Kunkle

Grenier-Boley

Sims

, et al. Genetic meta-analysis of diagnosed Alzheimer's disease identifies new risk loci and implicates Aβ, tau, immunity and lipid processing. Nat Genet 2019; 51: 414–430.

21.

Kurki

Karjalainen

Palta

, et al. Finngen provides genetic insights from a well-phenotyped isolated population. Nature 2023; 613: 508–518.

22.

Verbanck

Chen

Neale

, et al. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet 2018; 50: 693–698.

23.

Lin

Deng

Pan

. Combining the strengths of inverse-variance weighting and Egger regression in Mendelian randomization using a mixture of regressions model. PLoS Genet 2021; 17: e1009922.

24.

Carter

Sanderson

Hammerton

, et al. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol 2021; 36: 465–478.

25.

Tofighi

MacKinnon

. RMediation: an R package for mediation analysis confidence intervals. Behav Res Methods 2011; 43: 692–700.

26.

Bowden

Davey Smith

Burgess

. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol 2015; 44: 512–525.

27.

Bowden

Davey Smith

Haycock

, et al. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol 2016; 40: 304–314.

28.

Zheng

Baird

Borges

, et al. Recent developments in Mendelian randomization studies. Curr Epidemiol Rep 2017; 4: 330–345.

29.

Hartwig

Davey Smith

Bowden

. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol 2017; 46: 1985–1998.

30.

Pierce

Ahsan

Vanderweele

. Power and instrument strength requirements for Mendelian randomization studies using multiple genetic variants. Int J Epidemiol 2011; 40: 740–752.

31.

Burgess

Davey Smith

Davies

, et al. Guidelines for performing Mendelian randomization investigations: update for summer 2023. Wellcome Open Res 2019; 4: 186.

32.

Tsai

Chen

Chiou

, et al. Emergence of SGLT2 inhibitors as powerful antioxidants in human diseases. Antioxidants (Basel) 2021; 10: 1166.

33.

Pawlos

Broncel

Woźniak

, et al. Neuroprotective effect of SGLT2 inhibitors. Molecules 2021; 26: 7213.

34.

Kaneto

Obata

Kimura

, et al. Unexpected pleiotropic effects of SGLT2 inhibitors: pearls and pitfalls of this novel antidiabetic class. Int J Mol Sci 2021; 22: 3062.

35.

Hierro-Bujalance

Infante-Garcia

Del Marco

, et al. Empagliflozin reduces vascular damage and cognitive impairment in a mixed murine model of Alzheimer's disease and type 2 diabetes. Alzheimers Res Ther 2020; 12: 40.

36.

Iskander

Wang

, et al. Association of sodium-glucose cotransporter 2 inhibitors with time to dementia: a population-based cohort study. Diabetes Care 2023; 46: 297–304.

37.

Low

Goh

, et al. Association between use of sodium-glucose co-transporter-2 (SGLT2) inhibitors and cognitive function in a longitudinal study of patients with type 2 diabetes. J Alzheimers Dis 2022; 87: 635–642.

38.

Kellar

Craft

. Brain insulin resistance in Alzheimer's disease and related disorders: mechanisms and therapeutic approaches. Lancet Neurol 2020; 19: 758–766.

39.

Butterfield

Halliwell

. Oxidative stress, dysfunctional glucose metabolism and Alzheimer disease. Nat Rev Neurosci 2019; 20: 148–160.

40.

Miklankova

Markova

Hüttl

, et al. Empagliflozin alters lipid metabolism in the myocardium and liver in a prediabetes model with severe dyslipidemia. Front Pharmacol 2024; 15: 1393946.

41.

Mendez Espinoza

Choos

END

Ecelbarger

, et al. SGLT2 Inhibition leads to a restoration of hepatic and circulating metabolites involved in the folate cycle and pyrimidine biosynthesis. Am J Physiol Gastrointest Liver Physiol 2024; 327: G235–G253.

42.

Kastelein

van der Steeg

Holme

, et al. Lipids, apolipoproteins, and their ratios in relation to cardiovascular events with statin treatment. Circulation 2008; 117: 3002–3009.

43.

Ferrannini

Mark

Mayoux

. CV Protection in the EMPA-REG OUTCOME trial: a “thrifty substrate” hypothesis. Diabetes Care 2016; 39: 1108–1114.

44.

Würtz

Mäkinen

Soininen

, et al. Metabolic signatures of insulin resistance in 7,098 young adults. Diabetes 2012; 61: 1372–1380.

45.

Abdul-Ghani

DeFronzo

Norton

. Novel hypothesis to explain why SGLT2 inhibitors inhibit only 30–50% of filtered glucose load in humans. Diabetes 2013; 62: 3324–3328.

46.

Huang

Zhang

Yang

, et al. Circulating metabolites and risk of incident dementia: a prospective cohort study. J Neurochem 2023; 167: 668–679.

47.

Chhimpa

Singh

Puri

, et al. The novel role of mitochondrial citrate synthase and citrate in the pathophysiology of Alzheimer's disease. J Alzheimers Dis 2023; 94: S453–S472.

48.

Chen

Xiao

Liu

, et al. Dapagliflozin ameliorates ovulation disorders via attenuating activated microglia-mediated hypothalamic inflammation in HFD-fed mice. Neuroendocrinology 2024; 114: 331–347.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.24 MB