Hypocalcemia is a distinctive biochemical feature of hospitalized COVID-19 patients

Hypocalcemia has been recently identified as one of the major biochemical features of COVID-19 patients [1–7].

To date, several studies reported high frequency of hypocalcemia with both low ionized and total calcium levels in different cohorts of COVID-19 patients in emergency departments (ED) and during hospitalization. In some instances, also severe cases of hypocalcemia were reported [1–7]. Moreover, a strong association between lower calcium and higher levels of inflammatory parameters with increased disease severity has also been reported [1]. Finally, some authors identified hypocalcemia as a relevant and independent risk factor for worse clinical outcome including rate of hospitalization, ICU admission, and mortality [1, 5, 6].

In a previous study, in patients positive for nasopharyngeal swabs SARS-CoV-2 real-time reverse-transcriptase polymerase chain reaction (RT-qPCR) test, lower total and ionized calcium levels as compared to those admitted in ED with same clinical signs and symptoms and negative to RT-qPCR test were reported [4]. However, no clinical, biochemical, and disease severity parameters were reported for these two groups making difficult an accurate comparison.

The aim of our study was to compare, for the first time in the literature, ionized calcium levels between two groups of patients hospitalized in the same period of the year with severe acute respiratory infections COVID-19 and non-COVID-19 related.

Methods

We conducted a retrospective cohort study at IRCCS San Raffaele Hospital, a tertiary health-care hospital in Milan, Italy. We included patients (aged ≥ 18 years) with acute respiratory infectious diseases related to COVID-19 (CoV) and not related to COVID-19 (nCoV) hospitalized during the same study period (March–June 2020). We excluded patients transferred from other hospitals and patients initially hospitalized for other diseases.

Confirmed COVID-19 was defined as positive RT-qPCR from a nasal and/or throat swab together with signs, symptoms, and radiological findings suggestive of COVID-19 pneumonia.

Patients initially identified as negative for COVID-19 and successively positive at RT-qPCR or patients initially hospitalized for nCoV and then superinfected with SARS-CoV-2 were excluded from the study.

Only patients with serum ionized calcium levels data from arterial blood gas tests performed at initial evaluation in the ED were included. Ionized calcium levels were expressed both as actually measured levels (“actual calcium” AC) and as adjusted mathematically to a standardized pH of 7.4 levels (“standardized calcium” SC) to avoid influence of sample handling. Hypocalcemia was defined as calcium level below 1.18 mmol/L (RapidPoint 500 Analyzer, VA, USA).

We excluded patients with comorbidities and concomitant therapies influencing calcium metabolism: chronic kidney disease, osteoporosis, patients on glucocorticoids and antiepileptic drugs, vitamin D/calcium, loop/thiazide diuretics, and patients with an estimated glomerular filtration rate (eGFR) ≤ 30 mL/min/1.73 m² using creatinine levels at initial evaluation.

This study is part of the COVID-19 Biobank study, which is registered with ClinicalTrials.gov, NCT04318366 and obtained specific approval by the local EC [8].

Statistical analysis was conducted with SPSS 23.0 version (Chicago, USA). Categorical variables were indicated as frequency (%) and continuous variables as medians (IQR).

Differences in variable frequencies between groups were calculated using the Fisher test and the Mann–Whitney U test for continuous variables. All statistical tests were two sided. A p value of < 0.05 was considered statistically significant.

Results

A total of 20 non-COVID-19 patients were included in the study. These patients were matched with 20 COVID-19 patients (part of a previously reported cohort [1]) for the following known prognostic factors: age, sex, and concomitant comorbidities distribution.

Demographic and clinical patient characteristics are summarized in Table 1. Median [IQR] age was 72.5 years [68.2–86] in nCoV patients and 73.5 years [67.5–83.2] in CoV patients (p = 0.93) (Table 1). Fifty percent of patients in both groups were males (p = 1) (Table 1). The main comorbidities were history of arterial hypertension (70% in nCoV and 65% in CoV, p = 0.73), followed by diabetes mellitus (25% in nCoV and 35% in CoV, p = 0.73), coronary artery disease (25% in nCoV and 20% in CoV, p = 1), chronic obstructive pulmonary disease (20% in nCoV and 15% in CoV, p = 1), and active neoplasms (15% in nCoV and 15% in CoV, p = 1) (Table 1).

Table 1.

Comparison of clinical and biochemical parameters between non-COVID and COVID hospitalized patients

	Non-COVID (n.20)	COVID (n.20)
Clinical and biochemical variables			p value
Age (years)	72.5 [68.2–86]	73.5 [67.5–83.2]	0.93
Male gender (%)	10 (50)	10 (50)	1
BMI	26.7 [23.3–30.6]	28.4 [18.7–30]	0.56
Hypertension (%)	14 (70)	13 (65)	0.73
Coronary artery disease (%)	5 (25)	4 (20)	1
COPD (%)	4 (20)	3 (15)	1
Diabetes (%)	5 (25)	7 (35)	0.73
Cancera (%)	3 (15)	3 (15)	1
CRP (mg/L)	113 [50–157]	52 [28.5–141]	0.24
LDH (U/L)	295 [228.5–431]	357 [244–488]	0.39
Serum lactate (mmol/L)	1.05 [0.9–1.67]	1.25 [0.99–1.87]	0.29
Creatinine (mg/dL)	0.89 [0.64–1.18]	0.91 [0.78-1.26]	0.57
eGFR (mL/min/1.73 m2)	74 [49–93]	69 [56–84]	0.63
Tympanic temperature (°C)	37 [36.2–37.7]	37.4 [36.6–37.9]	0.44
Arterial pH	7.46 [7.42–7.5]	7.46 [7.44–7.48]	0.76
paO2/fiO2 ratio	323.6 [271.2–404.8]	277.4 [233.3–308.3]	0.041
satO2/fiO2 ratio	450 [428.6–458.3]	433.3 [363.9–442.8]	0.007
AC ionized calcium (mmol/L)	1.18 [1.12–1.2]	1.11 [1.06–1.16]	0.02
AC Hypocalcemia (%)	8 (40)	16 (80)	0.022
SC ionized calcium (mmol/L)	1.19 [1.14–1.22]	1.13 [1.09–1.17]	0.006
SC hypocalcemia (%)	7 (35)	15 (75)	0.025

^aOnly active neoplasms were included in this study.

BMI body mass index, COPD chronic obstructive pulmonary disease, CRP C-reactive protein, LDH lactate dehydrogenase, eGFR estimated glomerular filtration rate, calculated using the CKD-EPI equation, AC actual ionized calcium levels, SC standardized ionized calcium levels

Continuous variables are reported as medians (IQR)

P values reported in bold are those statistically significant

No statistical differences were found between the two groups regarding the inflammatory and organ injury parameters levels, as for serum creatinine and eGFR (Table 1).

Patients affected by COVID-19 presented a worse respiratory distress with a lower PaO2/FiO2 and SpO2/FiO2 ratios (calculated as the ratio between the arterial partial pressure or saturation of oxygen measured on arterial blood gas test and the fraction of inspired oxygen) compared to non-COVID-19 patients (Table 1).

A total of eight (40%) and seven (35%) non-COVID-19 patients were hypocalcemic using actual ionized calcium (AC) and standardized ionized calcium (SC) levels, respectively (Table 1). A total of 16 (80%) and 15 (75%) COVID-19 patients were hypocalcemic using AC and SC, respectively (Table 1). Compared to nCoV patients, hypocalcemia in CoV cohort was found statistically more frequent (p = 0.022 using AC; p = 0.025 using SC) (Table 1).

Lower calcium levels were statistically found in CoV patients compared to nCoV group (AC: 1.18 mmol/L [1.12–1.2] in nCoV vs. 1.11 mmol/L [1.06–1.16] in CoV, p = 0.02; SC: 1.19 mmol/L [1.14–1.22] in nCoV vs. 1.13 mmol/L [1.09–1.17] in CoV, p = 0.006) (Table 1).

On admission in ED, 15 patients (37.5%) (7 in CoV and 8 in nCoV group) had an eGFR from 30 to 60 mL/min/1.73 m². No statistical differences were found between patients with eGFR lower and higher than 60 mL/min regarding ionized calcium levels (AC: 1.13 mmol/L [1.1–1.17] in <60 mL/min group vs. 1.17 mmol/L [1.1–1.2] in ≥60 mL/min, p = 0.26; SC: 1.15 mmol/L [1.1–1.18] in <60 mL/min group vs. 1.18 mmol/L [1.12–1.21] in ≥60 mL/min, p = 0.33).

Linear regression analyses showed no statistical significant correlations of calcium levels with PaO2/FiO2 and SpO2/FiO2 ratios in CoV patients (PaO2/FiO2: AC p = 0.2, SC p = 0.13; SpO2/FiO2: AC p = 0.35, SC p = 0.29) and no significant differences were found between hypocalcemic and normocalcemic patients in both nCoV and CoV groups regarding these two ratios, although a tendency to a higher impact of hypocalcemia was noted in COVID-19 patients. (Table 2).

Table 2.

Comparison of PaO2/FiO2 and SpO2/FiO2 ratios between hypocalcemic and normocalcemic patients in COVID and non-COVID groups

Panel a	Hypocalcemic (AC)	Normocalcemic (AC)	p value
PaO2/FiO2 ratio
COVID	266 [214–303]	304 [282–366]	0.14
Non-COVID	354 [246–398]	309 [276–410]	0.63
SpO2/FiO2 ratio
COVID	419 [348–442]	442 [442–453]	0.08
Non-COVID	450 [417–457]	450 [439–461]	0.51

Panel b	Hypocalcemic (SC)	Normocalcemic (SC)	p value
PaO2/FiO2 ratio
COVID	257 [208–304]	300 [285–347]	0.13
Non-COVID	328 [238–404]	318 [278–404]	0.54
SpO2/FiO2 ratio
COVID	419 [346–442]	442 [442–452]	0.06
Non-COVID	447 [414–457]	452 [442–462]	0.33

Discussion

To the best of our knowledge, this is the first study comparing the occurrence of hypocalcemia and calcium levels in two matched groups of hospitalized patients for COVID-19 and non-COVID-19 infectious respiratory diseases. We found a higher rate of hypocalcemia with lower calcium levels in CoV compared to nCoV patients.

Rate of hypocalcemia occurring in our CoV study population was similar to other epidemiological data reported in previous studies, which found an, at least initially, unexpected high frequency of hypocalcemia [1–7].

Hypocalcemia was an already reported finding in patients hospitalized in Internal Medicine departments [9] and was found quite consistently to be associated with increased risk of acute respiratory failure and mortality [10, 11]. Therefore, it is currently unclear if hypocalcemia in COVID-19 is only a marker of disease severity rather than a specific feature of the disease. This open issue also impacts on the clinical and prognostic significance of this biochemical finding.

In a previous study, lower calcium levels were reported in patients positive to SARS-CoV-2 RT-qPCR test as compared to negative patients, hypothesizing a direct influence of viral infection on calcium metabolism [4]. However, no comparative clinical data were reported and thus also in this case a possible influence of differences in disease severity on calcium levels could not be excluded.

Therefore, in order to understand if hypocalcemia may be a distinctive feature of COVID-19 we thought of interest to compare calcium levels in hospitalized CoV and nCoV patients matched for main anthropometric, biochemical, and clinical features.

Inflammatory parameters, as CRP and LDH levels, were found to be negatively associated with calcium levels and, in different cohorts of COVID-19 patients, hypocalcemia and lower calcium levels resulted as an independent risk factor for worse clinical outcome, including hospitalization, ICU admission, and mortality. In order to reduce possible confounding influences of known COVID-19 prognostic factors as age, gender, and presence of concomitant comorbidities [12], we compared calcium levels in patients admitted to our ED for acute respiratory illness during the same period of time with or without SARS-CoV-2 infection matched for the above characteristics on an one case-one control basis.

Despite the same baseline clinical characteristics and inflammatory parameters of the two groups, we found a much higher rate of hypocalcemia with lower calcium levels in CoV patients compared to nCoV. Confirming the epidemiological data on SARS-CoV-2 infection, CoV patients were characterized by a worse respiratory distress compared to nCoV [12].

Our data confirm previous studies that reported hypocalcemia in hospitalized patients with acute illness [13]. However, the rate of hypocalcemia was doubled in COVID-19 patients matched for baseline anthropometric and clinical features. Therefore, the very high frequency of hypocalcemia seems to be a distinctive feature of SARS-CoV infections [14] and in particular of COVID-19. Pathophysiologically, it can be hypothesized that specific viral mechanisms that influence calcium handling [15–17] in the presence of widespread vitamin D deficiency [18–22] may contribute to specifically lowering calcium in the disease. Clinically, it may be hypothesized that low calcium levels may be one neglected determinant, as well as hypovitaminosis D [23, 24], of COVID-19 respiratory complications and mortality.

Therefore, serum calcium (and vitamin D) may be thought to be potential target for intervention in hospitalized COVID-19 patients who have been recently reported to be at high risk of vertebral fractures [7, 25–27]. Interestingly, we have recently reported that women with osteoporosis treated with vitamin D did not seem to be at increased risk of severe COVID-19 despite treatment with anti-osteoporotic drugs potentially predisposing to both infections and hypocalcemia [28].

One main limitation of our study is the limited number of patients with nCoV respiratory infection admitted to ED. However, it should be underlined that this is a remarkable cohort given the overwhelming number of COVID-19 patients hospitalized in that same study period. Other limitations of our report include the lack of outcome data, of clinical features possibly related to hypocalcemia, of calcium and vitamin D data during hospitalization as well as information on calcium (and vitamin D) administration in these patients.

In conclusion, our study suggests that hypocalcemia may be a distinctive biochemical feature of COVID-19 potentially impacting on disease clinical severity and representing a novel possible treatment target worth to be tested in this clinical setting.

Author contributions

All authors contributed equally.

Data availability

All authors had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent

Written informed consent was waived.