What are the mediators linking cardiometabolic risk factors and severe COVID-19?

By Neha MathurReviewed by Neha MathurAug 2 2023

In a recent article published in JAMA Network Open, researchers performed a genome-wide association study (GWAS) to explore whether particular biomarker levels are causally related to coronavirus disease 2019 (COVID-19) hospitalization.

They used a two-sample Mendelian randomization (MR) approach to analyze 235 protein biomarkers possibly associated with COVID-19 severity in people at higher risk of cardiometabolic disorders.

Background

Type 2 diabetes (T2D), obesity, and hypertension are some of the cardiometabolic parameters that increase the likelihood of severe COVID-19. Yet, studies have barely explored the biological mechanisms connecting cardiometabolic risk factors and serious COVID-19.

MR is a well-recognized statistical method to infer causality between exposure and outcome using their genetic associations. A previous MR-approach-based study identified angiotensin-converting enzyme 2 (ACE2) as a causal biomarker of COVID-19 severity.

Thus, it seems reasonable to take the MR approach and investigate which circulating proteins associate cardiometabolic parameters with COVID-19 severity, which might facilitate the development of novel therapies for COVID-19.

About the study

In the present study, researchers first identified genetic variants within 300 kilobases of the biomarker locus, cis-protein quantitative trait loci [pQTLs] correlated to biomarker levels in people with dysglycemia and other cardiovascular issues to explore their causal association with adverse COVID-19 prognosis, like COVID-19-related hospitalization. 

They validated their findings in three large independent populations. The first population was from the large Prospective Urban and Rural Epidemiological (PURE) biomarker project.

The other 2 study populations were from Folkersen et al. (n = 3,394 participants) and Sun et al. (n = 3,301 healthy adults of European ancestry), for which biomarker pQTL data were publicly available. The researchers also used genetic and proteomic data of 4,147 Outcome Reduction With Initial Glargine Intervention (ORIGIN) trial participants. 

They used a factorial design to randomly assign participants to insulin glargine versus standard care and an ω-3 fatty acid supplement versus placebo. They followed cardiovascular events and general health outcomes, on average, for 6.2 years. 

Further, the team assayed biomarker concentrations within a multiplex biomarker panel that if not normally distributed, were natural log–transformed to estimate a normal distribution. Eventually, they standardized values to a mean (SD) of zero.

The team used linear regression analyses to estimate the relationship between single-nucleotide variants (SNVs) and all biomarkers tested in this study. The main study outcome was hospitalization after confirmed COVID-19, as assessed via publically available data released by the COVID-19 Host Genetics Initiative. 

Results

A total of 238 biomarkers from 4,147 participants were deemed suitable for analysis. However, they could investigate only 235 using MR. Of the 235 biomarkers analyzed from the ORIGIN data, 15 showed a nominal association with COVID-19 hospitalization.

Of those, hepatitis A virus cellular receptor 1, also known as kidney injury molecule-1 (KIM-1), showed the highest association, such that its higher levels decreased the risk of COVID-19 hospitalization. 

KIM-1 membrane protein is expressed in the kidney, lung, liver, and spleen and regulates viral infection and autoimmunity through multiple mechanisms. KIM-1 is also an established blood and urine marker of acute kidney injury, with rising serum KIM-1 levels showing an association with glomerular filtration rate decline.

Previously reported clinical and epidemiologic association between KIM-1 and severe COVID-19 is consistent with the current study findings. 

With little data on the role of KIM-1 in COVID-19 severity, further work could confirm these findings. Nonetheless, the directionality of the observed association was consistent across all three independent cohorts and further reinstated a protective relationship between circulating KIM-1 and COVID-19 severity.

Furthermore, the KIM-1 concentration surged by 0.17 SD for every 1-kg/m2 increase in BMI, with no hint of directional pleiotropy. The multi-variate MR analysis indicated that KIM-1 also affected this association. So, while BMI was causally associated with severe COVID-19, increasing KIM-1 levels decreased the association of increased BMI with COVID-19 severity by ~10%.

Conclusions

The current MR analyses identified a protective association between KIM-1 levels and severe COVID-19, while BMI upregulated KIM-1 levels. Furthermore, MR analyses suggested that KIM-1 attenuated the association between BMI and COVID-19 hospitalization, although these findings warrant further investigation.

Intriguingly, KIM-1 work goes beyond regulating kidney function; accordingly, in this study, the researchers found a biological pathway linking BMI, KIM-1, and COVID-19 severity.