The COVID-19 pandemic, caused by the SARS-CoV-2 virus, has presented significant public health challenges worldwide ¹. Understanding host-pathogen interactions and immune responses triggered by SARS-CoV-2 infection is crucial for developing effective diagnostic and therapeutic strategies². Prior studies have highlighted the potential of blood metal levels as biomarkers for various diseases, including viral infections ^3,4. In a recent publication⁵, we reported potential correlations between blood metal levels in two groups of plasma samples. The first and second groups consisted of donors who had tested positive for SARS-CoV-2, confirmed via Reverse Transcription Polymerase Chain Reaction (RT-PCR), and of pre-pandemic donors who had never been exposed to the virus respectively. The first group included individuals with varying levels of disease severity. In a recent study ⁶, we used Logistic Regression (LR) on Laser-Induced Breakdown Spectroscopy (LIBS) data to demonstrate that sodium (Na), potassium (K), magnesium (Mg), and calcium (Ca) levels could differentiate between healthy individuals and SARS-CoV-2-infected patients. Our findings indicated that the combined analysis of the blood metals K, Na, and Mg was more effective in distinguishing between healthy and infected plasma samples than any single metal alone. Additionally, we observed a significant decrease in the Na/K LIBS emission line ratio, approximately doubling in COVID-19-positive samples compared to negative ones. These results suggest that subtle interactions among Na, K, Mg, and Ca in COVID-19 patients may contribute to biochemical alterations following SARS-CoV-2 infection.

Immunoglobulin G (IgG) is the predominant antibody subclass in the human circulatory system and plays a critical role in mediating adaptive immune responses. It is generated by the immune system in response to antigens, including those derived from pathogens such as SARS-CoV-2, the causative agent of COVID-19⁷. In the case of COVID-19, IgG antibodies typically become detectable in the blood several days to weeks following the onset of infection and can remain present for months^8,9. ​The enduring presence of IgG antibodies renders them to be essential indicators for detecting previous infections and verifying immune responses post-vaccination. Serological assays that identify IgG antibodies are extensively utilized to determine prior exposure to SARS-CoV-2, assess community-level immunity, and evaluate vaccine effectiveness¹⁰.

In this work, we build upon our previous LIBS studies⁶ to investigate the potential relationship between blood metal levels, and their ratios, and immune response to COVID-19 in a total of 150 plasma samples. The main difference with our previous study is that in this one, we compared the metal levels of 46 pre- and 64 post-pandemic donors as well as those of 40 donors whose COVID was not disclosed to the researchers conducting the spectroscopic analyses with IgG levels in their plasma samples. Given that COVID-19-positive donors experienced varying degrees of disease severity, their IgG levels also varied. Our objective was to determine whether changes in IgG levels correlated with variations in blood metal concentrations. We note that, in this study, we measured the LIBS elemental emission line intensities and their ratios. As all of the emission lines considered in this study exhibit no sign of self-absorption, we expect a linear relationship between emission intensity and concentration. As a result, LIBS-derived normalized intensity ratios can provide information on elemental concentration ratios.

LIBS is a rapid and versatile analytical technique that simultaneously determines the multi-elemental composition of samples in various states, with minimal or no sample preparation ^11–13. LIBS has emerged as a promising tool for real-time tissue diagnostics, allowing direct measurement of atomic emission spectra from biological samples ^14,15. Unlike conventional methods, LIBS enables rapid, label-free analysis of elemental composition, making it attractive in many applications^16,17 including for detecting electrolyte imbalances associated with disease states^4,18,19. However, certain limitations, such as self-absorption effects and spectral interference caused by the Stark effect, pose challenges to its clinical application ^15,20.

In this study, we used the same data acquired using the experimental setup described in our previous publications ^4,6. Therefore, we recall only the key components relevant to the current study.

Blood was collected from 150 donors aged 18–65 in the Boston and Florida regions. The samples were collected under the National Expanded Access protocol sponsored by the Mayo Clinic. Informed consent was obtained from all subjects and studies were performed in accordance with FDA guidelines and regulations including IRB approval. Sample collection was approved by the MRN Diagnostics Institutional Review Board. Plasma samples were collected in sodium citrate anticoagulant by MRN Diagnostics from 46 donors both before (pre-November 2019), 104 after the emergence of SARS-CoV-2 in the United States, but prior to the availability of COVID-19 vaccines. Consequently, none of the individuals who contributed samples for this study had been vaccinated against SARS-CoV-2. Blood samples were processed within 30 minutes of collection, with plasma separated and initially stored at -20°C before being transferred to -80°C. In the following sections, we use the terms Negative, Positive, and Blind in place of pre-pandemic, post-pandemic, and samples with unknown COVID-19 status, respectively. The Blind category refers to 40 blood plasma samples whose COVID-19 status was unknown to the researchers conducting the experiments. Plasma samples were centrifuged and deposited onto a silicon (Si) wafer before being dried under an infrared lamp for 10 minutes. To reduce measurement bias, all samples, regardless of COVID-19 status, were placed randomly on the Si wafer before acquiring laser-induced breakdown spectra.

We used a Q-switched Nd:YAG laser (Continuum Surelite II) to generate laser-induced plasma in the blood plasma samples. The laser delivered an energy of 130 ± 2 mJ per pulse, with a pulse duration of 7 ns. The beam was focused to a spot size of approximately 200 µm on the surface of the blood plasma samples using an air-spaced doublet lens with a focal length of 30 mm. LIBS experimental parameters were optimized to minimize self-absorption and saturation effects in the emission line intensities of Na, K, Mg and Ca. Spectral data were collected using an Echelle spectrograph coupled with an Intensified Charge-Coupled Device (ICCD) camera, acquiring 60 spectra per sample. Visual inspection of these spectra revealed initial low-intensity shots in each 15-shot set, likely due to surface contamination, which were excluded from analysis. The remaining 56 spectra were evaluated using a Z-score threshold of 1.0, retaining between 30 and 51 spectra per patient. These were used to compute a mean spectrum for each patient, subsequently normalized by dividing its intensity by the total sum of the measured spectral intensities.

To compare the LIBS spectra of negative, positive, and blind plasma samples, we calculated the area under the curve (AUC) for specific emission lines of key elements: Ca (Ca I 422.7 nm; Ca II 315.9, 393.4, and 396.8 nm), K (K I 766.5 and 769.9 nm), Mg (Mg I 285.2 nm; Mg II 280.3 nm), and Na (Na I 330.3 and 589.3 nm) across the remaining spectra. These values correspond to the normalized mean integrated line intensities. To better characterize the behavior of these elements across the negative, blind, and positive sample types, we calculated the average integrated intensity for each element (Ca, K, Mg, and Na) by averaging the normalized mean integrated line intensities of their respective emission lines. Hereafter, we refer to the average integrated intensity of each element simply as its "normalized intensity".

In addition to recording the LIBS spectra, we measured the IgG levels for all the 150 blood samples analyzed. IgG antibodies play a crucial role in the immune response to SARS-CoV-2 infection, and their presence typically indicates prior exposure or an ongoing immune response ²¹. Measuring IgG levels alongside elemental analysis allows for a somewhat more comprehensive comparison between SARS-CoV-2 infected patients and healthy controls than if each technique were used independently. It also provides an opportunity to better assess the potential of LIBS as a tool for blood analysis, offering insights into its potential to detect disease-related changes.

IgG antibody levels in plasma samples were measured using a commercial SARS-CoV-2 IgG chemiluminescent microparticle assay (LumiraDX, Abbott Industries), which targets antibodies against the receptor-binding domain (RBD) of the virus ²². The RBD is a key region of the SARS-CoV-2 spike protein responsible for binding to the Angiotensin-Converting Enzyme 2 (ACE2) receptor on host cells, facilitating viral entry. It is a major target of neutralizing antibodies, making it an important marker for assessing immune response. Using the LumiraDZ instrument, we measured antibody levels by taking the average signal of two channels that detect antibodies targeting the virus's receptor-binding domain (RBD).

Figure 1 illustrates the IgG antibody levels in plasma samples collected from negative, positive and blind plasma samples. It shows that negative plasma samples exhibit lower IgG levels than the positive and blind plasma samples. IgG values for negative samples range from 0.04 to 0.14, while those for blind and positive samples are above 0.15 and 2.65, respectively. This indicates no overlap in IgG values between negative samples and either blind or positive samples.

To explore the interplay between elemental composition and IgG levels, we have examined the relationship between IgG antibody levels and individual elemental concentrations, as well as their corresponding ratios and assessed potential correlations among these variables.

Figure 2 shows the relationship between IgG antibody levels and the normalized LIBS intensities of Ca, K, Na, and Mg for the negative, positive, and blind samples. For these three sample groups, the mean values and standard deviations were calculated. The results show that mean normalized elemental intensities in the positive and blind sample groups are comparable, with substantial overlap between them. In contrast, the negative samples generally exhibit lower mean values and less overlap with the other two groups. Among the elements analyzed, potassium (K) levels differ significantly between the negative samples and the other groups, with no observed overlap between them. Figure 2 shows also that for IgG levels close to zero, the concentrations of K, Na, Ca, and Mg are lower compared to samples with higher IgG levels. The findings might point toward a possible link between increased elemental concentrations and a more pronounced immune response.

We further examined the association between IgG antibody levels and elemental ratios. In Fig. 3, we present the results for three ratios of elements; additional ratios are reported in the Supplementary Materials (Figure S1). The findings are consistent with those observed for individual elemental concentrations: the positive and blind samples show similar mean values with considerable overlap, whereas the negative samples generally exhibit less overlap compared to the other two groups. Notably, the Na/K ratio stands out, showing a statistically significant difference between the negative group and both the positive and blind groups.

In Fig. 3 (top-left panel), the full range of IgG antibody levels is plotted against the normalized intensity of Na/K ratio using LIBS emission spectra. The bottom-left panel offers a magnified view of the transition zone between negative and positive samples, focusing on IgG antibody levels below 2.5. This shows that, as IgG levels change, the Na/K ratio follows a distinct trend, particularly in the transition region, suggesting that it may serve as a valuable marker for differentiating between negative and positive plasma samples. We observed in Fig. 3a significant decrease in the plasma Na/K ratio in individuals with detectable IgG antibody levels compared to negative samples. This decrease hints at a potential link between the Na/K ratio and the levels of IgG generated by the immune system in response to antigens associated with the SARS-CoV-2 virus. A similar pattern was observed for other metal ratios, such as Ca/K, suggesting a potential broader systemic reaction associated with the presence of IgG’s.

It is worth noting that, relative to the negative group samples, mean values of K in positive samples increased more (by about 2) than Na. Although K alone shows the greatest absolute shift, the Na/K ratio is a more robust marker owing to its lower betweensample dispersion, which yields narrower confidence intervals. Collectively, these observations suggest that elevated potassium contributes to the Na/K shift, while the ratio itself could provide a cleaner and more consistent metric for distinguishing early SARS-CoV-2 infection from negative controls.

The observed shifts in blood metal ratios, particularly the Na/K ratio, during SARS-CoV-2 infection may point to potential disruptions in cellular homeostasis and biochemical pathways. The decline in Na/K ratios across various stages of infection could suggest its utility as a biomarker for tracking disease progression and immune response. Further investigation will be important to clarify the underlying mechanisms driving these changes.

Figure 3 reveals the presence of two, and possibly three, distinct regions in the behavior of the plasma Na/K ratio as a function of IgG antibody levels. It is important to note that IgG values below 0.14 are considered indistinguishable from zero, as this range falls within the background signal of the assay. Thus, measurements below 0.14 reflect baseline noise rather than true IgG levels. The first region corresponds to individuals who had not been infected with SARS-CoV-2. As expected, within the sensitivity limits of the LumiraDX detection system their IgG levels are effectively zero, (but in practice, ranging from approximately 0.04 to 0.14). In this region, the Na/K ratio remains consistently around 1.1 ± 0.4, with values ranging approximately from 0.75 to 1.50. The second region includes individuals who had confirmed SARS-CoV-2 infections and exhibit IgG levels ranging from about 0.15 to 30. In this group, the Na/K ratio drops sharply to approximately 0.5 as soon as IgG levels become detectable and remain at this reduced level even as IgG levels increase further. One may also envision a transition region, where the Na/K ratio decreases rapidly from around 1.5 (characteristic of uninfected individuals) to about 0.5 (typical of infected individuals with detectable IgG). However, given the abruptness of this transition, it is not treated here as a separate region.

Next, we applied Pearson correlation analysis in an attempt to identify potential relationships between IgG antibody levels and both elemental intensities and their derived ratios²³. Linear associations are frequently used in biomedical research due to their clinical relevance and interpretability. In Fig. 4, we show the Pearson correlation coefficients for elements, the ratios of these elements, and IgG antibody levels. To account for baseline IgG levels observed in individuals never infected with SARS-CoV-2, we assigned a value of zero (0) for the IgG for these individuals and subtracted a value of 0.09, the average of the 0.04 to 0.14 baseline observed, from all IgG measurements of the positive patients. This correction is sufficient as the IgG for the positive plasma samples are at least one order of magnitude larger than the baseline. The heatmaps for the blind and positive samples were nearly identical, to facilitate comparison with the negative samples, only one is presented.

The heatmap reveals notable differences between the negative and positive groups. In the negative group, potassium (K) shows strong positive correlations with calcium (Ca) and magnesium (Mg) (Pearson coefficients of 0.80 and 0.81, respectively), whereas these correlations are substantially lower in the positive group (0.17 and 0.15). Similarly, sodium (Na) is highly correlated with Ca and Mg in the negative group (Pearson coefficients of 0.85 and 0.79) compared to almost negligible correlations in the positive group (0.03 and 0.05). Moreover, among the various elemental ratios examined, the largest discrepancy in Pearson correlation between the two types of samples is observed for the ratios Na/K and Mg/Na, with the negative samples exhibiting a coefficient of 0.32 and the positive ones a coefficient of -0.12.

Figure 4 shows that the Pearson correlation coefficients between elemental intensities or their ratios and IgG antibody levels are generally below 0.29, indicating these two parameters are not linearly correlated.

Several researchers have reported that elemental concentrations and their ratios play a crucial role in medical diagnostics ^24,25. Our results suggest that potassium exhibited a more pronounced difference between negative and positive samples, while sodium showed only a slight variation. This led to a normalized Na/K intensity ratio in negative samples that was approximately twice as high as in positive samples. This observation suggests that the relative concentration of sodium to potassium is nearly twofold higher in negative samples, hinting at the potential use of LIBS intensity ratios to approximate elemental concentration shifts and serve as diagnostically relevant markers.

The Na/K ratio is a critical indicator of electrolyte balance and plays a central role in essential physiological processes such as fluid homeostasis, nerve transmission, and muscle function ^26,27,28. Sodium contributes to blood pressure regulation through its effects on fluid balance and vascular volume, while potassium counterbalances sodium by promoting vasodilation and lowering blood pressure. An altered Na/K ratio has been linked to increased risks of hypertension, stroke, and cardiovascular disease ^26,29. The kidneys maintain this delicate balance by regulating the excretion of both electrolytes³⁰. We note that SARS-CoV-2 infection has been shown to disrupt electrolyte homeostasis, including the Na/K ratio, through multiple mechanisms ³¹. The virus can affect kidney function directly via viral entry into renal cells or indirectly through systemic inflammation and cytokine release potentially altering sodium and potassium handling. Furthermore, angiotensin-converting enzyme 2 (ACE2), the receptor used by SARS-CoV-2 to enter host cells, is expressed in renal and cardiovascular tissues, suggesting a pathway through which the virus can influence fluid and electrolyte regulation. Changes in the Na/K ratio observed in patients infected with the SARS-CoV-2 virus may reflect both direct viral effects and systemic responses, such as immune activation and vascular stress. These imbalances might serve as potential biomarkers of disease severity and complications associated with SARS-CoV-2 infection. Given the possible clinical relevance of Na/K imbalances, additional studies will be important to clarify their role in disease progression and to evaluate the feasibility of LIBS-based techniques for monitoring electrolyte variations in patient populations.

This study has several limitations. First, the analysis was based on a dataset of blood samples obtained from 150 individual patients. Although this number provides valuable initial insights, statistical tests such as the t-test, which we employed, generally require larger sample sizes to ensure greater statistical power and robustness of the conclusions. Second, the study included a set of blinded samples intended for validation; however, the majority of these were subsequently identified as positive cases. This outcome raises concerns about the appropriateness of treating these samples as truly blind, and one could argue for their incorporation into the positive cohort. We chose not to do so, as this would have resulted in an unbalanced dataset that would have conflicted with the pre-defined research design. A third limitation arises from the lack of comprehensive clinical metadata such as disease severity, treatment regimens, comorbidities, and other relevant medical factors. These data could have enriched the interpretation of our findings. The absence of longitudinal follow-up data and the inability to obtain serial samples from the same individuals limited our capacity to explore temporal dynamics of metal and IgG antibodies with disease progression. Addressing these limitations in future studies will be useful to expand upon our current findings.

Our study offers preliminary insights into changes in blood metal ratios associated with SARS-CoV-2 infection and the immune response. We have observed that when IgG antibodies become detectable, the plasma sodium-to-potassium (Na/K) ratio drops by approximately 50% relative to SARS-CoV-2-negative controls. This decrease is sustained even as IgG concentrations continue to rise, possibly indicating a persistent shift in electrolyte balance concurrent with the developing adaptive immune response. Other elemental ratios, such as Ca/K and K/Mg, also varied systematically with IgG levels, although these changes were less pronounced than the Na/K shift. The present study has several limitations. First, the relatively small sample size constrains the generalizability of the findings. Second, our analysis of the LIBS data assumed that the measured spectral intensities vary linearly with the element concentrations within the samples. Although we have not observed any evidence of self-absorption in the LIBS emission lines, other plasma mechanisms may have influenced the measured intensity and as a result may not accurately reflect true concentrations. These factors, combined with inherent LIBS constraints (such as spectral interference and shot-to-shot variability), could have introduced additional uncertainty to the overall uncertainty budget and influence the quantitative interpretation of the results. We suspect that such mechanisms may have less impact when comparing ratios of emission lines and IgG concentrations. Third, the study is based on a single time-point measurements per individual and therefore precludes conclusions about the temporal progression of these elemental changes.

Taken together, these observations tentatively suggest that the onset of an IgG immune response may be accompanied by alterations in the blood’s metal profile. To our knowledge, this is among the first studies to explore a possible link between LIBS spectral signatures and serological status in the context of viral infection. It provides a perspective that contributes to gaining insights related to physiological disturbances accompanying the onset of the immune response, pointing to a potential bridge between atomic spectroscopy and immunodiagnostics. The approach used in this study, comparing LIBS spectral signatures to antibody levels following viral infection, needs further investigations and could, with additional validation, help inform the development of new diagnostic strategies.