The coronavirus disease 2019 (COVID-19) is a global pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which has rapidly spread across the world (1). To date, the global count of confirmed COVID-19 cases has surpassed 676.5 million, with over 6.88 million deaths reported (source: https://coronavirus.jhu.edu/map.html). Patients who have recovered from COVID-19 are gradually reporting various complications, such as fatigue, a loss of the sense of smell or taste, impaired pulmonary function, neurological diseases and bone loss (2, 3). Therefore, considering the high infectivity of SARS-CoV-2 and severe adverse consequences of COVID-19, it is necessary to understand the causal relationship between COVID-19 and complications and to take effective preventive measures.

Recent study findings suggest that there is a close correlation between COVID-19 and osteoporosis (OP) (4). OP is a metabolic bone disease that is characterized by a reduction in bone mineral density (BMD), which can lead to bone fragility and an increased risk of fractures (5, 6). However, current research related to COVID-19 and OP is mainly focused on how to prevent and treat OP in patients with COVID-19 (7, 8), and there is a lack of large prospective cohort studies or randomized controlled trials (RCTs) to evaluate the effect of COVID-19 on OP. In addition, RCTs for COVID-19 are difficult to conduct due to the need for extensive human resources and time-consuming follow-up. Furthermore, existing observational studies may have biased conclusions due to the possibility of confounders. To minimize the impact of confounders on the association between COVID-19 and OP, a more efficient method for inferring potential causal relationships is needed.

In recent years, Mendelian randomization (MR) analysis has been widely used in causal inference in epidemiology (9, 10). The main principle behind this approach involves utilizing genetic variants as instrumental variables (IVs) to establish a causal relationship between exposure and outcome (11). As shown in Figure 1, the IVs in an MR analysis should satisfy the three core assumptions of relevance, independence and exclusivity (12); that is, (1) the IVs should be strongly correlated with exposure, (2) the IVs should be independent of confounders that affect the exposure-outcome relationship, and (3) the IVs should be capable of affecting the outcome only through exposure and have no direct correlation with the outcome. Ran et al. conducted a two-sample MR analysis for the association of total body BMD with severe COVID-19 (13). Their results indicate that BMD might be a useful predictor of severe COVID-19 in older adult populations and could help identify individuals at higher risk of disease progression. However, it did not delve into the impact of COVID-19 on osteoporosis.

As there is currently no conclusive proof to establish a causal relationship between COVID-19 and OP, we used single nucleotide polymorphisms (SNPs) strongly associated with COVID-19 severity (SARS-CoV-2 infection, COVID-19 hospitalization, and severe COVID-19) as IVs, and we carried out a two-sample MR analysis using genome-wide association study (GWAS) summary statistics to explore the causal effect of COVID-19 severity on OP.

There is a lack of large prospective cohort studies or RCTs to evaluate the effect of COVID-19 on OP. Furthermore, existing studies may have biased conclusions due to the possibility of confounders. To minimize the impact of confounders on the association between COVID-19 and OP, we used SNPs strongly associated with COVID-19 severity (SARS-CoV-2 infection, COVID-19 hospitalization, and severe COVID-19) as IVs and carried out a two-sample MR analysis using GWAS summary statistics, and we obtained more reliable results. Because genetic variants are formed in utero and remain with the person throughout life, exposure differences throughout life exist between genetic subgroups, and because an exposure precedes an outcome, thus being temporally sequential, reverse causality is less likely. Additionally, the examination of each SNP as an IV revealed that it was associated with the exposure factor COVID-19 but not with other phenotypes, and it is unlikely that it would affect the outcome through other genetic pathways. Identifying the causal relationship between COVID-19 and OP is important for the prevention, diagnosis and treatment of OP. To our knowledge, this is the first two-sample MR study to explore the causal effect of COVID-19 severity on OP.

There could be various factors that potentially explain the link between COVID-19 and OP. Current research suggests that SARS-CoV-2 may affect bone homeostasis by directly or indirectly affecting osteoblasts and osteoclasts, leading to bone loss. The effects of SARS-CoV-2 on humans have been found to be attributable to its binding to the angiotensin-converting enzyme 2 receptor (ACE2) (24). SARS-CoV-2 downregulates ACE2 expression and enhances angiotensin II (Ang II) levels upon the infection of target cells (25). Shimizu et al. demonstrated that Ang II significantly induced the expression of receptor activator of NF-kappaB ligand (RANKL) in osteoblasts, leading to the activation of osteoclasts (26). A cytokine storm might be another important cause of abnormal bone metabolism. Upon SARS-CoV-2 infection, innate immune cells such as macrophages and neutrophils are immediately activated, leading to a limitation of infection. This activation triggers downstream activation of the persistent adaptive immune system, resulting in the production of neutralizing antibodies and T-cell responses against the virus. However, when inflammation serves its purpose and is not resolved, it leads to dysregulated hyperinflammation, a cytokine storm, and the suppression of the adaptive immune system, which further escalates tissue damage and organ failure. A cytokine storm is characterized by the uncontrolled production of multiple inflammatory cytokines, including IL-6, IL-7, IL-2, IL-17, TNF-α, as well as monocyte chemoattractant protein (MCP)-1 and macrophage inhibitory protein (MIP)-1α (27, 28). Qiao et al. demonstrated that bone loss is associated with SARS-CoV-2-induced cytokine dysregulation, as circulating proinflammatory cytokines not only upregulate osteoclastic differentiation in bone tissues but also trigger an amplified proinflammatory cascade in skeletal tissues to augment their pro-osteoclastogenesis effect (29). It has been suggested that hypoxia and oxidative stress may play a role in the development of osteoporosis in COVID-19 patients (30). Severe COVID-19-induced hypoxemia can trigger the excessive production of reactive oxygen species (ROS), which disrupts the balance of redox homeostasis (31). This disruption has been found to induce apoptosis of osteoblasts and osteoclasts, and to regulate the expression of RANKL/osteoprotegerin (OPG), leading to the generation of osteoclasts and ultimately bone loss (32). Glucocorticoids are beneficial in the treatment of acute respiratory distress syndrome (ARDS) by reducing inflammation and improving the function of the lungs and extrapulmonary organs and are therefore widely used in patients with COVID-19 in most parts of the world (33). However, it is considered to be a double-edged sword in the treatment of patients with COVID-19. Glucocorticoids affect bone homeostasis by inhibiting osteoblast osteogenesis and promoting osteoclast resorption, leading to bone loss, and have been found to be the cause of medically induced OP, commonly known as glucocorticoid-induced osteoporosis (GIOP) (34). Therefore, there is a link and interaction among COVID-19, glucocorticoids and osteoporosis that deserves the attention of clinicians and researchers. Another possible reason is that with the COVID-19 pandemic, many countries implemented an unprecedented array of measures to mitigate the spread of the virus, including mass social isolation, travel bans, restrictions on public gatherings, and national lockdowns (35). While these social distancing strategies were necessary from a public health perspective, they presented challenges in the management of many chronic diseases. A study found that physical activity decreased significantly during the COVID-19 pandemic (36), which undoubtedly had an impact on the onset or progression of OP.

In contrast, there was no evidence observed in our two-sample MR analysis to support the association of genetically predicted OP with COVID-19 severity in individuals of European descent based on the results. Therefore, it is suggested that COVID-19 patients may not require special preventive or treatment measures for osteoporosis. However, we cannot exclude the possibility that we failed to detect the association due to the limitation of this study. First, this study used a population sample of European origin and lacked data from other ethnic groups, so the extrapolation of the results is limited, and data from other ethnic groups are needed for analysis and comparison to make the results more reliable. Second, this study used summary GWAS data and was unable to assess the nonlinear relationship between exposure and outcome, and because of the lack of individual data, it was not possible to stratify the analysis by sex, age, or the site of OP. In addition, there were too few SNPs as IVs in this study, which might have some influence on the results, especially for R² and MR power. If a larger sample could be obtained and more SNPs could be extracted, the reliability of the results could be further verified. Furthermore, the severity of COVID-19 is likely to be impacted by a myriad of factors, including the healthcare infrastructure, social contact patterns, environmental conditions, and viral strain mutations, among others. It is important to note that these factors can interact with one another, making it challenging to isolate their individual impacts on the severity of COVID-19. Additionally, these factors are complex and challenging to explain through MR Analysis.

In summary, the results of the MR analysis provide preliminary evidence that a genetic causal link between the severity of COVID-19 and OP may be absent. Therefore, it is suggested that COVID-19 patients may not require special preventive or treatment measures for osteoporosis. However, the contribution of other factors cannot be dismissed. To corroborate the study’s conclusions, additional MR analyses incorporating more extensive GWAS summary data and a larger set of genetic instruments, coupled with sizable prospective cohort studies or RCTs, are indispensable.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

KZ as the first author performed data analysis and wrote the manuscript. HL, QX, ZJL, WC, DL, and XW contributed suggestions for manuscript revision and revised the manuscript. WS, XZ, RP, XL, ZJ, ZXL, and CX provided advice and suggestions while we met some problems during the data analysis process. HZ and ZJL conceived and initiated this project, provided advice on experimental design, oversaw the implementation of the statistical method, and revised or finalized the manuscript. All authors contributed to the article and approved the submitted version.

WC was employed by R&D Center, Youjia (Hangzhou) Biomedical Technology Co., Ltd.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.