Skip to main content


  • Research
  • Open Access

Serum magnesium levels and lung cancer risk: a meta-analysis

  • 1,
  • 1Email author,
  • 2,
  • 3 and
  • 2
World Journal of Surgical Oncology201816:137

  • Received: 25 January 2018
  • Accepted: 5 July 2018
  • Published:



Whether serum magnesium levels were lower in patients with lung cancer than that in healthy controls is controversial. The aim of this study was to identify and synthesize all citations evaluating the relationship between serum magnesium levels and lung cancer.


We searched PubMed, WanFang, China National Knowledge Internet (CNKI), and SinoMed databases for relevant studies before December 31, 2017. Two authors independently selected studies, extracted data, and assessed risk of bias.


Eleven citations comprising 707 cases with lung cancer and 7595 healthy controls were included in our study. Serum magnesium levels were not significantly lower in patients with lung cancer [summary SMD = 0.193, 95%CI = − 1.504 to 1.890] when compared to health controls, with significant heterogeneity (I2 = 99.6%, P < 0.001) found. Negative associations were found among Asian populations [summary SMD = 0.229, 95%CI = − 1.637 to 2.094] and European populations [summary SMD = − 0.168, 95%CI = − 0.482 to 0.147]. No publication bias was found using the test of Egger and funnel plot.


Our study suggested that serum magnesium levels had no significant association on lung cancer risk.


  • Magnesium level
  • Lung cancer
  • Meta-analysis
  • Healthy controls


Lung cancer is the leading cause of death from cancer, resulting 1.38 million people deaths each year [1]. Its 5-year survival rate is still as low as 15%, and it is poor while compared with those in high incidence of other cancer [2]. Previous studies pointed out that lung cancer is the most common cancer among men and women, and both developed and developing countries bear a huge social and economic burden [3]. Previous publications proved that both genetic and environment factors were related to lung cancer risk [47]. Furthermore, trace-heavy elements also played a significant role on human health and disease [8, 9], as well as lung cancer [10].

Magnesium is one of the trace elements in our bodies, and to date, some papers had been published to investigate the association between serum magnesium levels and lung cancer risks. Two papers [11, 12] reported a higher of serum magnesium level in cases with lung cancer, while six papers [1318] found a lack of significant association. Conversely, three papers [1921] suggested that it is lower in lung cancer cases when compared to the healthy controls. Therefore, the aim of this study was to identify and synthesize all citations evaluating the relationship between serum magnesium levels and lung cancer risk.


Study selection

A comprehensive literature search was conducted in platforms of PubMed, WanFang, China National Knowledge Internet (CNKI), and SinoMed databases up to December 31, 2017. Free words adopted were as follows: “magnesium” or “Mg” combined with “lung cancer” or “lung carcinoma” without restrictions. Reference lists of the studies retrieved were also examined to find any additional study potentially unidentified. The course of study selection was completed by two investigators independently. Any resulting discrepancies were resolved by a third reviewer.

The inclusion criteria were as follows: (i) having a prospective design or a case-control design or a cross-sectional study; (ii) evaluating the association between serum magnesium levels and risk of lung cancer; (iii) reporting mean and standard deviation (SD) of magnesium levels (or sufficient data to compute them) both in lung cancer patients and healthy controls; and (iv) studies published in English language or Chinese language. If more than one article referred to the same populations, only the study that included the most lung cancer cases or the latest publication was included.

Data extraction and quality assessment of studies

Two investigators independently extracted the following data: (1) first author’s last name; (2) publication year; (3) study design; (4) country; (5) number of lung cancer cases and participants; (6) sex of cases; (7) age range or mean age of the cases; (8) mean and SD of magnesium levels both in lung cancer patients and healthy controls; and (9) method used for detection of magnesium. Any resulting discrepancies were resolved by a third reviewer.

The methodological quality of studies was evaluated independently by two researchers using the Newcastle-Ottawa Quality Assessment Scale [22]. The three components were as follows: (1) patient selection (4 points); (2) comparability (2 points); and (3) outcome (3 points) for a total score of 9 points.

Statistical analysis

Standardized mean difference (SMD) and their 95% confidence interval (CI) were calculated for relationship between serum magnesium levels and lung cancer risk. A random effect model was used in our meta-analysis [23]. The heterogeneity among studies was evaluated with I2 and Q tests. [24]. P < 0.05 in Q test and I2 > 50% indicated statistically significant heterogeneity [25]. Meta-regression was adopted to assess the between-study heterogeneity. Egger’s regression asymmetry test [26] and funnel plot [27] were used to visually examine publication bias on study outcome. Statistical analyses were performed using STATA version 12.0 (Stata Corporation, College Station, TX, USA). A two-sided P < 0.05 was defined as statistical significance.


Study characteristics

As shown in Fig. 1, the initial 486 articles screened through databases of PubMed, WanFang, China National Knowledge Internet (CNKI), and SinoMed databases searching and 1 additional record identified through other sources. There are 372 articles that were reviewed in the title and abstract while excluding the duplications from different databases. Three hundred and forty two of 372 articles were excepted when screened on the basis of title and abstract; 30 articles were examining full texts. Nineteen studies were further excluded (reviews, not report mean or SD, animal studies, letter to the editors). Finally, 11 articles [1121] were eligible to be included in the analysis comprising 707 patients with lung cancer and 7595 healthy controls. All the included studies were case-control studies. Nine studies were carried out from China, 1 from Spain, and 1 from Turkey. Ten of the included studies used the methods of atomic absorption spectrophotometer measurements for detection of magnesium. In the study quality assessment, all the included studies were with a score greater or equal to 6. The basic features of all citations are shown in Table 1.
Fig. 1
Fig. 1

Flowchart of meta-analysis for exclusion/inclusion of studies

Table 1

Characteristics of all included studies

Study, year


Age (range or Mean ± SD)

Study type

Lung cancer cases


Methods of measured magnesium


Female (%)

Magnesium: Mean ± SD


Magnesium: Mean ± SD

Cobanoglu U et al., 2010


54 ± 8.29




156.21 ± 22.21 μg/L


185.8 ± 4.05 μg/L

Atomic Absorption Spectrophotometer measurements (UNICAM-929 spectrophotometer)

Diez M et al., 1989


60 ± 7




20.6 ± 3.2 μg/L


21.7 ± 8 μg/L

Perkin-Elmer 5.000 atomic absorption spectrophotometer

Jin ZJ et al., 2001






1300 ± 390 μmol/L


1320 ± 310 μmol/L

Atomic Absorption Spectrophotometer measurements

Xu ZF et al., 1993


56 ± 7.5




804.63 ± 71.29 μmol/L


936.83 ± 93.31 μmol/L

Atomic Absorption Spectrophotometer measurements

He WD et al., 1995






940.88 ± 116.95 μmol/L


871.24 ± 96.88 μmol/L

Atomic Absorption Spectrophotometer measurements

Huang ZY et al., 1998






1.8275 ± 0.375 μmol/L


0.8254 ± 0.1778 μmol/L

Atomic Absorption Spectrophotometer measurements (Japan Shimadzu-AA670/C2H2)

Wang ZL et al., 2003






68.29 ± 35.26 μg/L


114.1 ± 52.12 μg/L

Atomic Absorption Spectrophotometer measurements and 721 spectrophotometer

Du FL et al., 1996






1100 ± 300 μmol/L


1100 ± 100 μmol/L

Atomic Absorption Spectrophotometer measurements

Guo XH et al., 1994






20.88 ± 6.72 μg/mL


18.84 ± 5.86 μg/mL

Atomic Absorption Spectrophotometer measurements (Varian Spectr AA-40p, USA)

Wang FJ et al., 2014






880 ± 60 μmol/L


860 ± 90 μmol/L

Xylene blue method

Fang JQ et al., 1998






1.34 ± 0.35 μmol/L


1.36 ± 0.29 μmol/L

Atomic Absorption Spectrophotometer measurements

Serum magnesium levels and lung cancer risk

Pooled results suggested that magnesium levels in patients with lung cancer was not significantly lower than healthy controls [summary SMD = 0.193, 95%CI = − 1.504 to 1.890, I2 = 99.6%, Pfor heterogeneity < 0.001] (Fig. 2). When we performed the subgroup analysis by geographic location, the association was not significant either in Asian populations [summary SMD = 0.229, 95%CI = − 1.637 to 2.094] or in European populations [summary SMD = − 0.168, 95%CI = − 0.482 to 0.147].
Fig. 2
Fig. 2

The forest plot of the relationship between serum magnesium levels and lung cancer risk

Sources of heterogeneity and meta-regression

Meanwhile, I2 was 99.6% (p < 0.001) for the pooled sensitivity, suggesting high heterogeneity in the sample of studies. Univariate meta-regression was then carried out to determine the reason of heterogeneity. However, there were no significant contributions about publication year, case number, geographic location, sex, and different methods on this high between-study heterogeneity.

Sensitivity analysis and publication bias

Sensitivity analysis conducted while removing one study at the time revealed that no single study had essential effect on the whole result. Figure 3 showed that no publication was considered by the funnel plot method on the basis of data, as well as the Egger’s test (P = 0.586).
Fig. 3
Fig. 3

Funnel plot for the analysis of publication bias between serum magnesium levels and lung cancer risk


In this study, we assessed the association between serum magnesium levels and risk of lung cancer. We did not find a positive association between serum magnesium levels and lung cancer risk. Through our subgroup analysis, we further found no significant association among Asian and European populations. Significant heterogeneity between studies observed in this meta-analysis should be considered as a major limitation of these findings; however, heterogeneity was mainly related to strength of the association rather than the direction of risk estimate, suggesting overall promising findings on the outcome investigated in the present study.

Two previous prospective cohort studies concluded that higher category of dietary magnesium intake had no significant association on lung cancer risk among German population and China population [28, 29]. However, a report [30] had been resulting that higher magnesium levels in drinking water could reduce the risk of lung cancer deaths in women. To our knowledge, no comprehensive analysis had been published to assess the serum magnesium levels on lung cancer risk. In our study, we did not find significant association of lower serum magnesium levels in patient with lung cancer. However, level of magnesium in other disease may be in the normal range, and that magnesium can have an effect on this disease [31].

The existence of heterogeneity among the studies, which is common in meta-analyses [32], may affect the pooled results. Meta-regression was performed to find the potential covariates (publication year, case number, geographic location, sex, and different methods to detect magnesium levels) which may cause this high heterogeneity. However, no covariate was found to significantly contribute to heterogeneity. In our study, most of the included studies obtained nonsignificant association between serum magnesium levels and lung cancer risk. Only one study [12] reported that serum magnesium level in patient with lung cancer is extremely higher than that in healthy controls. We reviewed the article again and confirmed the data exacted from the study; no error was made. Sensitivity analysis was performed, and no study had essential effect to the significant between-study heterogeneity and the whole result. On the other hand, we used a random effect model to combine the results. As we all know, random effect model had wider rage about 95%CI than fix effect model and could obtain more accurate results. Furthermore, only three studies [12, 13, 17] reported the types and staging of lung cancer, which may also be a factor on the between-study heterogeneity. Therefore, studies with detailed information of types and staging of lung cancer are wanted to further explore this association.

Some advantages existed in our study. Firstly, a comprehensive literature search was performed to investigate the relationship between serum magnesium levels and lung cancer risk. Secondly, most of the included studies involved large numbers of patients and healthy controls, and this may strengthen the power of the pooled results. Thirdly, there was no significant publication when tested by Egger and funnel plot, which indicates that our results are stable.

The present study has some limitations. Firstly, the individual studies may have failed to control for potential confounders, which may introduce bias in an unpredictable direction. Secondly, ten of 11 studies were from Asia, and 9 were from China, and thus, more related researches from other countries are wanted to verify the association between geographic location and lung cancer risk.


Based on the obtained results, we concluded that serum magnesium levels may have no significant association in patients with lung cancer. As we experienced some limitations in our study, such as more studies were from Asia, further studies are wanted to confirm this finding.



Confidence intervals


Standard deviation


Standard mean differences


Availability of data and materials

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Authors’ contributions

XNZ was the guarantor of integrity of the entire study and was responsible for the definition of intellectual content and manuscript editing. XHS contributed to the study concepts, study design, and data analysis and was responsible for the manuscript preparation. KJT and GW were responsible for the literature research. GW carried out the experimental studies. KJT and YJ were responsible for the data acquisition. All authors read and approved the final manuscript.

Competing interest

The authors declare that they have no competing interests.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

Department of respiration, the First Affiliated Hospital of Guangxi Medical University, N0.6 Shuangyong Road, Nanning, 530021, Guangxi, China
Department of rheumatism, Liuzhou Worker’s Hospital, Liuzhou, 545005, Guangxi, China
Department of neurosurgery, Liuzhou General Hospital, Liuzhou, 545006, Guangxi, China


  1. Torre LA, Bray F, Siegel RL, et al. Global cancer statistics, 2012. CA Cancer J Clin. 2015;65(2):87–108.View ArticlePubMedGoogle Scholar
  2. Tsao AS, Scagliotti GV, Bunn PA Jr, et al. Scientific advances in lung cancer 2015. J Thorac Oncol. 2016;11(5):613–38.View ArticlePubMedGoogle Scholar
  3. Minguet J, Smith KH, Bramlage P. Targeted therapies for treatment of non-small cell lung cancer—recent advances and future perspectives. Int J Cancer. 2016;138(11):2549–61.View ArticlePubMedGoogle Scholar
  4. Liu C, Cui H, Gu D, et al. Genetic polymorphisms and lung cancer risk: evidence from meta-analyses and genome-wide association studies. Lung Cancer. 2017;113:18–29.View ArticlePubMedGoogle Scholar
  5. Wang J, Liu Q, Yuan S, et al. Genetic predisposition to lung cancer: comprehensive literature integration, meta-analysis, and multiple evidence assessment of candidate-gene association studies. Sci Rep. 2017;7(1):8371.View ArticlePubMedPubMed CentralGoogle Scholar
  6. Papadopoulos D, Papadoudis A, Kiagia M, et al. Nonpharmacologic interventions for improving sleep disturbances in patients with lung Cancer: a systematic review and meta-analysis. J Pain Symptom Manag. 2018;55(5):1364–81.View ArticleGoogle Scholar
  7. Poinen-Rughooputh S, Rughooputh MS, Guo Y, et al. Occupational exposure to silica dust and risk of lung cancer: an updated meta-analysis of epidemiological studies. BMC Public Health. 2016;16(1):1137.View ArticlePubMedPubMed CentralGoogle Scholar
  8. Demir N, Basaranoglu M, Huyut Z, et al. The relationship between mother and infant plasma trace element and heavy metal levels and the risk of neural tube defect in infants. J Matern Fetal Neonatal Med. 2017. Epub ahead of print.Google Scholar
  9. Wu T, Bi X, Li Z, et al. Contaminations, sources, and health risks of trace metal (loid) s in street dust of a Small City impacted by artisanal Zn smelting activities. Int J Environ Res Public Health. 2017;14(9):961.View ArticlePubMed CentralGoogle Scholar
  10. Zablocka-Slowinska K, Placzkowska S, Prescha A, et al. Serum and whole blood Zn, Cu and Mn profiles and their relation to redox status in lung cancer patients. J Trace Elem Med Biol. 2018;45:78–84.View ArticlePubMedGoogle Scholar
  11. He WD. Detection and analysis of some trace elements in serum of patients with lung cancer. Jiujiang Med J. 1995;10(2):69–71.Google Scholar
  12. Huang ZY, Hu FD. Comparative study of serum trace elements in patients with lung cancer. Shanxi Clin Med J. 1998;17(2):114–6.Google Scholar
  13. Diez M, Arroyo M, Cerdan FJ, et al. Serum and tissue trace metal levels in lung cancer. Oncology. 1989;46(4):230–4.View ArticlePubMedGoogle Scholar
  14. Jin ZJ, Qian LQ, Dong GQ, et al. Measurement and analysis of serum copper, zinc and magnesium in patients with lung cancer and gastric cancer. Shaanxi Med J. 2001;30(3):165–6.Google Scholar
  15. Du FL, Li ZM, Cao MJ, et al. Determination of serum copper, zinc, magnesium and iron in patients with pulmonary tuberculosis, chronic bronchitis, pulmonary heart disease and lung cancer. J of Xi’an Med University. 1996;17(3):348–50.Google Scholar
  16. Guo XH, Li PF, Peng FK, et al. Relationship between serum zinc, copper, manganese and lung cancer. Chin Pub Heal. 1994;10(4):156–7.Google Scholar
  17. Wang FJ, Zhao ZE, Wen JB, et al. Levels of some major elements in patients with lung cancer: a retrospective analysis. Chin J Gen Pract. 2014;12(4):528–30.Google Scholar
  18. Fang JQ, Mei YL, Li LJ, et al. Analysis of selenium, copper, magnesium, magnesium and selenium contents in elderly patients with lung cancer. Ningxia Med J. 1998;20(4):247–8.Google Scholar
  19. Cobanoglu U, Demir H, Sayir F, et al. Some mineral, trace element and heavy metal concentrations in lung cancer. Asian Pacific journal of cancer prevention : APJCP. 2010;11(5):1383–8.PubMedGoogle Scholar
  20. Xu ZF, Sun YC, Zhang CW, et al. Clinical significance of changes of serum copper, zinc and magnesium contents in patients with lung cancer. J of Second Mil Med Univ. 1993;14(2):195–6.Google Scholar
  21. Wang ZL, Zhang W, Zhang HY, et al. Determinat ion of serum trace elements and their clinical value in patients with lung cancer. Clin Focus. 2003;18(4):183–5.Google Scholar
  22. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol. 2010;25(9):603–5.View ArticlePubMedGoogle Scholar
  23. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7(3):177–88.View ArticlePubMedGoogle Scholar
  24. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557–60.View ArticlePubMedPubMed CentralGoogle Scholar
  25. Higgins JP, Thompson SG. Controlling the risk of spurious findings from meta-regression. Stat Med. 2004;23(11):1663–82.View ArticlePubMedGoogle Scholar
  26. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–34.View ArticlePubMedPubMed CentralGoogle Scholar
  27. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4):1088–101.View ArticlePubMedGoogle Scholar
  28. Li K, Kaaks R, Linseisen J, et al. Dietary calcium and magnesium intake in relation to cancer incidence and mortality in a German prospective cohort (EPIC-Heidelberg). Cancer Causes Control. 2011;22(10):1375–82.View ArticlePubMedGoogle Scholar
  29. Takata Y, Shu XO, Yang G, et al. Calcium intake and lung cancer risk among female nonsmokers: a report from the Shanghai Women’s Health Study. Cancer Epidemiol Biomarkers Prev. 2013;22(1):50–7.View ArticlePubMedGoogle Scholar
  30. Cheng MH, Chiu HF, Tsai SS, et al. Calcium and magnesium in drinking-water and risk of death from lung cancer in women. Magnes Res. 2012;25(3):112–9.PubMedGoogle Scholar
  31. Srebro DP, Vučković SM, Dožić IS, et al. Magnesium sulfate reduces formalin-induced orofacial pain in rats with normal magnesium serum levels. Pharmacol Rep. 2018;70(1):81–6.View ArticlePubMedGoogle Scholar
  32. Munafo MR, Flint J. Meta-analysis of genetic association studies. Trends Genet. 2004;20(9):439–44.View ArticlePubMedGoogle Scholar


© The Author(s). 2018