Introduction

Calcium is the most abundant mineral in the human body, 99% of which is stored in bones and teeth1,2. Calcium homeostasis is critical for maintaining normal muscle contractility, vascular tone, and electrical signaling1. An increasing number of people take calcium supplements as they are able to prevent osteoporosis and fractures3. Recently, a rising body of studies has suggested that serum calcium levels are associated with cardiovascular disease4,5,6,7,8,9,10,11. In addition, a recent study showed that the intermountain risk score, which includes calcium, is independently related to both short- and long-term mortality among AMI patients12. However, the clear relationship between serum calcium and mortality in patients with AMI is controversial. Most previous studies have found that low serum calcium is associated with higher mortality in patients with AMI, whereas high serum calcium is not associated with higher mortality4,6,8,9,13. However, some recent studies have found that high serum calcium is also associated with higher mortality in AMI patients, and we found that these studies included more patients with hypercalcemia5,14. Therefore, we speculate that both high and low serum calcium might be associated with higher mortality in AMI patients. This study further explored the association between serum calcium and in-hospital mortality in a large cohort of patients with AMI.

Methods

Data source

The data of this study were extracted from the Philips eICU Collaborative Research Database. The database was a multi-center intensive care unit (ICU) database for over 200,000 admissions in 2014 and 2015 from 335 ICUs in 208 hospitals in the United States of America15. This database consists of vital sign measurements, care plans, the severity of illness measures, laboratory tests, diagnosis information, treatment information, and more. Before requesting access to the eICU Collaborative Research Database, you must complete the CITI “Data or Specimens Only Research” course. Then you can apply for registration to obtain permission to use the eICU Collaborative Database. After the registration application has been approved, the data will be disclosed. If the protocol is approved, the data can be directly downloaded from the eICU Collaborative Database (http://eicu-crd.mit.edu). The Massachusetts Institute of Technology (Cambridge, MA) approved this data collection and waived informed consent requirements. This study's first author (Dingfeng Fang) has approved access to the eICU Collaborative Database (certification number: 50,924,352). The authors are grateful to the original study group for providing data for the current analysis. The study report follows the broad EQUATOR guidelines16. All methods in our study were performed under the Declaration of Helsinki.

Study design and participants

This study is a retrospective cohort study based on the eICU Collaborative Research Database (N = 200,859). Patients with a diagnosis of AMI were included (N = 9884). All AMI patients were continuously enrolled unless: (1) serum calcium level was unrecorded; (2) albumin level was unrecorded. 7284 AMI patients were included in the final data analysis. The details of inclusions and exclusions were showed in Fig. 1.

Figure 1
figure 1

A flow chart of the inclusion and exclusion of patients.

Full size image

Outcome and data collection

The primary clinical outcome of this study was all-cause in-hospital death. To make the data more reliable, serum calcium (mg/ml) was corrected to albumin (g/dL) using the Payne formula: corrected serum calcium (corrected sCa) = Serum calcium + 0.8 × (4.0—albumin)17. Demographic data, medical history, laboratory tests, and treatment information were extracted from the database.

Statistical analyses

Each patient's corrected sCa was calculated and categorized into four groups: ≤ 8.5, 8.5–9.5, 9.5–10.5, and > 10.5 mg/dL. Normally distributed continuous variables were presented as mean ± SD. Skewed distributed continuous variables were presented as median [interquartile range]. The categorical variables were presented as numbers (n) and percentages (%). The characteristics of this study population according to the corrected serum calcium levels were compared using a Kruskal–Wallis test or one-way analysis of variance (ANOVA) for continuous variables and a Chi-Square test for categorical variables.

We performed univariate analysis to detect the possible risks associated with in-hospital death. Multiple logistic regression models were performed to evaluate the association between corrected calcium and in-hospital death. Multivariate adjusted models were applied as well as non-adjusted models. The adjusted confounders were selected basis on their association with the primary outcome. Different stratification was adjusted for different numbers of confounding variables. Furthermore, the threshold effect of corrected sCa on in-hospital death was explored using piecewise linear regression according to the smoothing plot.

A P value less than 0.05 was considered significant for all tests. Analysis was performed using the statistical software packages R (http://www.Rproject.org, The R Foundation) and Empower Stats (http://www.empowerstats.com, X&Y Solutions, Inc., Boston, MA).

Results

Baseline characteristics of the study population

As shown in Fig. 1, among the 200,859 patients from the eICU Collaborative Research Database, 193,575 patients were excluded based on exclusion criteria. A total of 7284 patients were eventually enrolled in our study, of which 799 (10.97%) died in the hospital. Patients were classified into four groups according to corrected sCa: Q1 (≤ 8.5 mg/dL), Q2 (8.5– ≤ 9.5 mg/dL), Q3 (9.5– ≤ 10.5 mg/dL), Q4 (> 10.5 mg/dL). Table 1 describes the baseline data for all patients and patients stratified by corrected sCa levels. Participants with higher corrected sCa levels were older, female predominant, and more likely to have atrial fibrillation, chronic renal insufficiency, COPD, and diabetes, while the patients with lower corrected sCa were more likely to have a cardiac arrest.

Table 1 Characteristics of study patients.
Full size table

The association between serum calcium and mortality

The details of univariate analysis in Table 2 showed that age, being female, BMI, hemoglobin, TC, LDL-C, HDL-C, history of atrial fibrillation, cardiac arrest, CRI, COPD, hypertension, diabetes mellitus, stroke or history of malignancy, and treatment with antiplatelet, nitroglycerin or mechanical ventilation were strongly correlated with the in-hospital death. Table 3 illustrated that multivariate analysis demonstrated that corrected sCa was an independent predictor of in-hospital death. In Model 1 (non-adjusted model), the risk of in-hospital mortality first decreased and then increased with increasing corrected sCa (categorical variable: Q2 vs. Q1, OR 0.5, 95% CI 0.4–0.7, P < 0.001; Q3 vs. Q1, OR 0.8, 95% CI 0.6–1.0, P = 0.035; Q4 vs. Q1, OR 1.6, 95% CI 1.1–2.3, P = 0.008). After adjusting for age, sex, BMI, and chronic renal insufficiency, the association between corrected sCa and in-hospital death was consistent with Model 1 (categorical variable: Q2 vs. Q1, OR 0.5, 95% CI 0.4–0.7, P < 0.001; Q3 vs. Q1, OR 0.7, 95% CI 0.5–0.9, P = 0.003; Q4 vs. Q1, OR 1.2, 95% CI 0.8–1.7, P = 0.425). Furthermore, this association remained stable when age, sex, BMI, chronic renal insufficiency, COPD, diabetes mellitus, stroke, history of malignancy, hemoglobin, TC, LCL-C, HDL-C, lipid-lowering drugs, antiplatelet, anticoagulants, and nitroglycerin were adjusted in Model 3 (categorical variable: Q2 vs. Q1, OR 0.4, 95% CI 0.2–0.7, P = 0.001; Q3 vs. Q1, OR 0.5, 95% CI 0.5–0.5, P = 0.009; Q4 vs. Q1, OR 0.9, 95% CI 0.4–2.0, P = 0.732).

Table 2 Univariate analysis.
Full size table
Table 3 Association of corrected serum calcium and the incidence of in-hospital mortality.
Full size table

Piecewise linear regression model

After the possible factors related to in-hospital death were adjusted, a U-shaped relationship between corrected serum calcium levels and in-hospital mortality in patients with AMI was observed (Table 4; Fig. 2). The occurrence of in-hospital death decreased with an increase in corrected sCa when the corrected sCa < 9.4 mg/dL (OR 0.8, 95% CI 0.7–0.9, P < 0.001). The occurrence of in-hospital death increased with an increase in corrected sCa when the corrected sCa > 9.4 mg/dL (OR 1.5, 95% CI 1.3–1.8, P < 0.001).

Table 4 Threshold effect analysis of serum calcium on in-hospital mortality.
Full size table
Figure 2
figure 2

The illustrated curved line relation between serum calcium and in-hospital mortality. (A) The density of the scatter points represents the number of patients. (B) Fitted curve was drawn to more clearly present the association between corrected serum calcium and in-hospital mortality in patients with AMI. According to the threshold analysis, the curve's inflection point is 9.4 mg/dL (2.35 mmol/L). The area between two blue lines (dotted lines) is a 95% CI.

Full size image

Discussion

This study retrospectively analyzed the association between corrected sCa levels and in-hospital mortality in patients with AMI. To analyze the in-hospital survival of 7284 patients with AMI, we found a U-shaped association between patients' corrected sCa levels on admission and in-hospital mortality. The results suggest that both high and low corrected sCa levels are associated with a higher risk of in-hospital mortality (Corrected sCa < 9.4 mg/dL, OR 0.8, 95% CI 0.7–0.9, P < 0.001; corrected sCa > 9.4 mg/dL, OR 1.5, 95% CI 1.3–1.8, P < 0.001).

Our findings are consistent with Shiyovich A et al.’s recent studies5 reported a significant U-shaped association between corrected sCa and in-hospital death, with corrected sCa ≤ 9.12 mg/dL and ≥ 9.86 mg/dL as independent predictors of significantly increased in-hospital mortality. Shiyovich A et al. retrospectively analyzed the relationship between corrected sCa and in-hospital mortality in 11,446 AMI patients and drew a broken line graph. In addition, multiple studies have found that high serum calcium levels are associated with higher in-hospital mortality10,11,14. Our study draws a precise smooth curve and identifies the inflection point of the curve (corrected sCa = 9.4 mg/dL), which further elucidates the relationship between serum calcium and in-hospital mortality in patients with AMI. A meta-analysis of prospective cohort studies18 found a U-shaped relationship between dietary calcium intake and cardiovascular mortality (Inflection point of calcium intake = 800 mg/day), which may correspond to our study.

The relationship between serum calcium and in-hospital mortality in patients with myocardial infarction remains controversial. In a 2021 study by Timo Schmitz et al., low serum calcium is associated with higher long-term mortality in patients with AMI, whereas high serum calcium was not4. Consistent with Timo Schmitz et al., multiple previous studies have not found that high serum calcium is associated with an increased risk of in-hospital mortality in patients with AMI8,9,13. An 8-year, single-center study in China6 found that higher serum calcium was associated with a lower risk of in-hospital death (N = 1431; Q1: < 8.56 mg/dL, Q2: 8.56–9 mg/dL, Q3: 9–9.44 mg/dL, Q4: ≥ 9.44 mg/dL). Thus, studies that are inconsistent with our study are not uncommon. We speculate that the main reasons for the different results are: (1) the study population is different; (2) the adjusted variables are different; (3) the statistical methods used are different. However, it is worth noting that our data include more hypercalcemia patients (Q4: corrected serum calcium ≥ 10.5 mg/dL, N = 296) because our study mainly focused on severe cases from the ICU of the United States of America. Therefore, our data may be more representative.

The mechanism responsible for the U-shaped relationship between serum calcium levels and in-hospital mortality in AMI patients is unclear. Several mechanisms may explain that low serum calcium is associated with higher in-hospital mortality in patients with AMI: (1) Serum calcium levels are directly related to the electrophysiological characteristic of the cardiomyocyte membrane. Low serum calcium may cause delayed closure of calcium channels, leading to prolonged plateaus13. In addition, patients with low serum calcium levels are more prone to ventricular arrhythmias and cardiac arrest19,20. (2) Low serum calcium is associated with multiple cardiovascular risk factors such as hypertension and left ventricular systolic dysfunction21,22. (3) Low calcium may lead to reversible heart failure and cardiomyopathy23,24. The mechanisms by which hypercalcemia leads to higher in-hospital mortality in patients with AMI may include the following: (1) High serum calcium levels may lead to subendothelial calcium deposition, leading to atherosclerosis1,25. (2) High serum calcium is associated with vascular calcification by binding to calcium-sensing receptors and inducing mineralization of smooth muscle cells5,26. (3) High serum calcium level increases the risk of heart failure, stroke, and type 2 diabetes mellitus10,21,27,28. One possibility deserves our attention: AMI leads to changes in serum calcium. However, this possibility is minimal because many prospective studies have found an association between serum calcium and in-hospital mortality in patients with AMI29,30,31.

Whether calcium supplementation can improve prognosis in patients with AMI remains controversial3,32,33,34,35. However, it is worth noting that research in recent years has shown that calcium intake from the diet can benefit cardiovascular health. A geospatial analysis found an inverse association between high calcium spring water and cardiovascular mortality, citing waters with high calcium content as cardiovascular protective36. Multiple studies on drinking water have found the same result37,38. A meta-analysis found a U-shaped relationship between dietary calcium intake and cardiovascular mortality, and calcium intake ≥ 800 mg/d is associated with increased cardiovascular risk. Therefore, supplementing calcium in the diet may be beneficial, but only in moderation. In contrast, calcium intake from calcium supplements was seemly associated with an increased risk of cardiovascular death25,39,40,41. An IVUS study by Najdat Bazarbashi et al. found that oral calcium supplements may increase calcium deposition in coronary vessels, leading to poor prognosis42. The mechanism of oral calcium supplements increasing cardiovascular risk is unclear, likely associated with the sudden increase in serum calcium. However, it is worth noting that the symptoms of severe AMI patients with low serum calcium significantly improved after calcium supplementation5. Based on these controversies, our study has great clinical value for managing patients with AMI. For the first time, our study demonstrated the U-shaped association between corrected serum calcium and in-hospital mortality in AMI patients with a smooth curve and identified the inflection point of the curve (corrected sCa = 9.4 mg/dL). Clinicians can refer to our findings to formulate calcium supplementation strategies for AMI patients, thereby reducing patient mortality.

This study has the following advantages: (1) The sample size is relatively large, and the serum calcium level of the patients was corrected using albumin, which makes the results more reliable; (2) The population of this study is critically ill patients from the ICU of the United States including more patients with abnormal serum calcium; therefore, our data are more representative; (3) This study is the first time to show the relationship between serum calcium and in-hospital death in AMI patients with a smooth curve and to determine the inflection point of the curve, which has an immense clinical value. However, this study excluded patients other than AMI and may not apply to all populations. In addition, due to the nature of observational studies, we can only observe associations, not causality. Our data come from critically ill patients in US ICUs with complex conditions. Although we have adjusted for confounders that may have been associated with the results, we cannot rule out some undocumented confounders confounding the results. Therefore, it is necessary to conduct higher-level clinical studies in more populations in the future to validate our findings.

Conclusions

The corrected serum calcium level is an independent prognostic marker of in-hospital mortality in patients with AMI with a U-shaped association (inflection point: corrected sCa = 9.4 mg/dL). It is necessary to carry out clinical studies with a higher level of evidence to validate our findings in the future.