Lack of significant association between dyslipidemia and COVID-19 mortality

Abstract

Recently, Aung et al. reported that distribution frequency of angiotensin converting enzyme (ACE) insertion/insertion (II) genotype had a significant impact on coronavirus disease 2019 (COVID-19) mortality.1Aung A.K. Aitken T. Teh B.M. Yu C. Ofori-Asenso R. Chin K.L. et al.Angiotensin converting enzyme genotypes and mortality from COVID-19: an ecological study.J. Infect. 2020; 81 (PubMed PMID: 33197472. Pubmed Central PMCID: 7666537): 961-965Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar Dyslipidemia is one of the most comorbidities among COVID-19 patients, however, the conclusions from published articles on the relationship between dyslipidemia and COVID-19 mortality are still controversial. For instance, several studies found that there was a significant relationship between dyslipidemia and an increased risk for mortality among COVID-19 patients,2Grasselli G. Greco M. Zanella A. Albano G. Antonelli M. Bellani G. et al.Risk factors associated with mortality among patients with COVID-19 in intensive care units in Lombardy, Italy.JAMA Intern. Med. 2020; 180 (PubMed PMID: 32667669. Pubmed Central PMCID: 7364371): 1345-1355Crossref PubMed Scopus (786) Google Scholar, 3Mayer M.A. Vidal-Alaball J. Puigdellivol-Sanchez A. Marin Gomez F.X. Leis A. Mendioroz Pena J Clinical characterization of patients with COVID-19 in primary care in catalonia: retrospective observational study.JMIR Public Health Surveill. 2021; 7 (PubMed PMID: 33496668. Pubmed Central PMCID: 7871981): e25452Crossref PubMed Scopus (8) Google Scholar, 4Rosenthal N. Cao Z. Gundrum J. Sianis J. Safo S. Risk factors associated with in-hospital mortality in a US national sample of patients with COVID-19.JAMA Netw. Open. 2020; 3 (PubMed PMID: 33301018. Pubmed Central PMCID: 7729428)e2029058Crossref PubMed Scopus (229) Google Scholar while other studies reported that dyslipidemia was not significantly associated with COVID-19 mortality.5An C. Lim H. Kim D.W. Chang J.H. Choi Y.J. Kim S.W. Machine learning prediction for mortality of patients diagnosed with COVID-19: a nationwide Korean cohort study.Sci. Rep. 2020; 10 (PubMed PMID: 33127965. Pubmed Central PMCID: 7599238): 18716Crossref PubMed Scopus (91) Google Scholar,6Graziani D. Soriano J.B. Del Rio-Bermudez C. Morena D. Diaz T. Castillo M. et al.Characteristics and prognosis of COVID-19 in patients with COPD.J. Clin. Med. 2020; 9 (PubMed PMID:33053774. Pubmed Central PMCID: 7600734)Crossref Scopus (29) Google Scholar Therefore, there is an urgent need to address the relationship between dyslipidemia and COVID-19 mortality by a quantitative meta-analysis. It has been reported that demographical characteristics (age and gender) and certain comorbidities (diabetes mellitus, cardiovascular disease, hypertension, chronic kidney disease and autoimmune diseases, etc.) are well-known modulators that affect the clinical outcomes of COVID-19 patients,7Biswas M. Rahaman S. Biswas T.K. Haque Z. Ibrahim B. Association of sex, age, and comorbidities with mortality in COVID-19 patients: a systematic review and meta-analysis.Intervirology. 2020; (PubMed PMID: 33296901): 1-12PubMed Google Scholar, 8Liang X. Shi L. Wang Y. Xiao W. Duan G. Yang H. et al.The association of hypertension with the severity and mortality of COVID-19 patients: evidence based on adjusted effect estimates.J. Infect. 2020; 81 (PubMed PMID: 32593655. Pubmed Central PMCID: 7315979 conflicts of interest): e44-ee7Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar, 9Yang H. Xu J. Liang X. Shi L. Wang Y. Autoimmune diseases are independently associated with COVID-19 severity: evidence based on adjusted effect estimates.J. Infect. 2020; (PubMed PMID: 33383087)https://doi.org/10.1016/j.jinf.2020.12.025Abstract Full Text Full Text PDF Scopus (13) Google Scholar suggesting that these factors might modulate the relationship between dyslipidemia and COVID-19 mortality. Thus, in this current meta-analysis, risk factors-adjusted effect estimates rather than crude effect estimates were utilized to calculate the pooled effect sizes. We did this systematic meta-analysis in accordance with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). All potentially eligible articles published between January 1, 2020 and February 26, 2021 were identify in the online databases (PubMed, Web of Science and EMBASE) with the following keywords: “SARS-CoV-2” or “severe acute respiratory syndrome coronavirus 2” or “COVID-19” or “coronavirus disease 2019” or “2019-nCoV” or "2019 novel coronavirus" and “dyslipidemia” or “hyperlipidemia” or “low-density lipoprotein” or “high-density lipoprotein” or “triglycerides” or “total cholesterol”. Reference lists of eligible articles were also searched to look for additional studies. The exposure group was defined as COVID-19 patients with dyslipidemia and the control group was defined as COVID-19 patients without dyslipidemia. The outcome of interest was mortality, which was defined as mortality, death, died, non-survivor, fatality or deceased. All peer-reviewed articles published in English language reporting the risk factors-adjusted effect estimate on the relationship between dyslipidemia and COVID-19 mortality were eligibly selected. Accordingly, we excluded preprints, case reports, review papers, corrections, comments, animal studies and in vitro studies, studies reporting crude effect estimate, studies without sufficient data and studies reporting clinical outcomes as severe/critical illness, intensive care unit admission, invasive mechanical ventilation/intubation or composite outcomes rather than mortality. Essential information including first author, number of COVID-19 patients, gender distribution, age (mean and standard deviation (SD) or median and interquartile range (IQR)), study design, region/country, clinical outcomes, adjusted effect estimates and adjusted variables was extracted from each included study (Table 1).Table 1Characteristics of the included studies.AuthorCountryCases (n)Male (%)Age (years)§The values are presented as mean ± standard deviation or median (interquartile range, IQR); USA, the United States of America; CI, confidence interval; OR, odds ratio; RR, risk ratio; RH, relative hazard; HR, hazard ratio.Study designAdjusted-effect (95% CI)Adjusted variablesClinical outcomesHashemi et al. (PMID: 32585065)USA36355.4%63.4 ± 16.5Retrospective studyOR: 0.91 (0.46–1.81)Chronic liver disease, age, obesity, male, cardiac diseases, hypertension, diabetes, pulmonary disordersDeathPettit et al. (PMID: 32589784)USA23847.5%58.5 ± 17Retrospective studyOR: 1.7 (0.4–7.0)Obesity, age, gender, hypertension, diabetes, pulmonary disease, cardiovascular disease, kidney disease, cancer, stroke, venous thromboembolismMortalityGrasselli et al. (PMID: 32667669)Italy398879.9%63 (56–69)Retrospective studyHR: 1.25 (1.02–1.52)Age, men, respiratory support, hypertension, heart disease, type 2 diabetes, malignancy, chronic obstructive pulmonary disease, angiotensin-converting enzyme inhibitor therapy, angiotensin receptor blocker therapy, statin, diuretic, positive end-expiratory pressure at admission, fraction of inspired oxygen at admission, arterial partial pressure of oxygen/fraction of inspired oxygen at admissionMortalityTartof et al. (PMID: 32783686)USA691645%49.1 ± 16.6Retrospective studyRR: 1.47 (1.02–2.11)Body mass index, age, sex, race/ethnicity, smoking, metastatic tumor/cancer, myocardial infarction, other immune condition, organ transplant, congestive heart failure, peripheral vascular disease, cerebrovascular disease, chronic pulmonary disease, renal disease, hypertension, asthma, diabetes mellitus status, timeDeathCzernichow et al. (PMID: 32815621)France579565.4%59±14Retrospective studyOR: 0.97 (0.74–1.27)Body mass index, age, diabetes, hypertension, sleep apnea, chronic kidney disease, heart failure, malignancies, history of smoking, sexMortalityNimkar et al. (PMID: 32838205)USA32755.7%71 (59–82)Retrospective studyOR: 1.4 (0.8–2.2)Race, chronic kidney disease in addition to six essential covariates (age, sex, race, hypertension, diabetes, cardiac disease)MortalityGiorgi-Rossi et al. (PMID: 32853230)Italy265350.1%72±24Prospective studyHR: 1.4 (0.9–2.2)Age, sexDeathEsme et al. (PMID: 32871002)Turkey16,94249%71.2 ± 8.5Retrospective studyOR: 0.77 (0.64–0.93)Gender, hypertension, diabetes mellitus, chronic obstructive pulmonary disease, coronary artery disease, atrial fibrillation, chronic kidney disease, dementia, depression, malnutritionMortalityYan et al. (PMID: 32949175)China110348.6%63 (51–71)Retrospective studyHR: 1.91 (0.46–7.99)Age, male, diabetes, hypertension, chronic obstructive pulmonary disease, chronic heart diseases, chronic kidney diseases, chronic liver diseases, cerebrovascular diseases, tumor, C-reactive protein, d-dimerMortalityIoannou et al. (PMID: 32965502)USA10,13191%63.6 ± 16.2Longitudinal corhot studyHR: 0.96 (0.83–1.11)All sociodemographic characteristics, comorbid conditions, symptomsMortalityGhany et al. (PMID: 33024960)USA40040%72±8Retrospective studyRH: 0.99 (0.99–1.00)Age, gender, charlson scoreDeathGraziani et al. (PMID: 33053774)Spain14,33949%66±15Retrospective studyOR: 1.03 (0.81–1.31)Chronic obstructive pulmonary disease, sex, age, heart failure, high blood pressure, stroke, arrythmia, ischemic heart disease, diabetes, sleep apnea, pulmonary thromboembolism, smokingDeathAn et al. (PMID: 33127965)Korea10,23739.9%44.97±19.79Nationwide cohort studyHR: 0.89 (0.66–1.20)Age, sex, income level, residence, household type, disability, symptom, infection routeDeathZhang et al. (PMID: 33122929)China9859.2%63.9 ± 1.4Retrospective studyOR: 2.94 (1.22–7.12)Age, gender, lymphocyte count, glycated hemoglobin, hypersensitive C-reactive protein, N-terminal brain natriuretic propeptide, creatinineMortalityShah et al. (PMID: 33169090)USA48756.1%68±17Retrospective studyOR: 1.36 (0.83–2.21)Age, gender, patient admitted from home, hypertension, cardiomyopathy, atrial fibrillation, chronic obstructive pulmonary disease, cerebrovascular accident, diabetes mellitus, acute kidney injury, initial chest x-ray/computed tomography findings, dyspnea in emergency department noted as positiveMortalityTomasoni et al. (PMID: 33179839)Italy69269.5%67.4 ± 13.2Retrospective studyHR: 0.82 (0.47–1.44)Age, sex, heart failure, hypertension, atrial fibrillation, coronary artery disease, chronic obstructive pulmonary disease, chronic kidney disease, oxygen saturation, arterial partial pressure of oxygen/fraction of inspired oxygen, hemoglobin, lymphocytes count, estimated glomerular filtration rate, C-reactive protein on admission, troponinDeathLoffi et al. (PMID: 33229434)Italy125263.7%64.7 ± 15.5Retrospective studyHR: 0.94 (0.63–1.41)Sex, left ventricular ejection fraction < 35%, cerebrovascular disease, atrial fibrillation, diabetes mellitus, hypertension, coronary artery disease, chronic kidney disease, ageDeathRossi et al. (PMID: 33222020)Italy59067.6%76.2 (68.2–82.6)Retrospective studyHR: 1.108 (0.859–1.431)Age, gender, temperature, arterial partial pressure of oxygen / fraction of inspired oxygen, lactate dehydrogenase, C-reactive protein, white blood cell count, lymphocytes rate, cardiovascular disease, diabetes, atrial fibrillation, chronic obstructive pulmonary disease, chronic kidney disease, stroke, malignancy, 3 or more comorbidities, angiotensin-converting enzyme inhibitor, angiotensin receptor blockers, calcium-channel blockers, alpha blockers, diuretics, beta blockersMortalityRosenthal et al. (PMID: 33301018)USA35,30253.4%63.6 ± 17.7Retrospective studyOR: 1.11 (1.03–1.19)Age, sex, race, payer type, admission point of origin, hospital region, hospital beds, hospital teaching status, statin, vitamin C, zinc, angiotensin-converting enzyme inhibitor, b blocker, calcium channel blocker, hydroxychloroquine and azithromycin use, sepsis, acute kidney failure, hypokalemia, hyperkalemia, hyponatremia, acidosis, acute liver damage, neurological disorder, myocardial infarction, congestive heart failure, cerebrovascular disease, chronic pulmonary disease, dementia, diabetes, any malignant neoplasm, metastatic solid tumor, hemiplegia, acquired immunodeficiency syndrome, hypertensionMortalityOzyilmaz et al. (PMID: 33322097)Turkey10572.4%45 (20–87)Retrospective studyOR: 4.060 (0.011–1555.792)Troponin I, C-reactive protein, lymphocyte count, shortness of breath, hypertension, diabetes mellitus, coronary artery diseaseMortalityLohia et al. (PMID: 33453090)USA187151.6%66 (54–75)Retrospective studyOR: 0.97 (0.76–1.23)Age, sex, race, smoking, body mass index, insurance and comorbidities which include coronary artery disease, congestive heart failure, chronic obstructive pulmonary disease, asthma, chronic kidney disease, end-stage renal disease on dialysis, any malignancy, any liver disease, history of previous stroke, hypertension, diabetesMortalityGupta et al. (PMID: 33461499)USA47345.5%70 (61–80)Retrospective studyOR: 1.28 (0.75–2.19)Race, age, sex, coronary artery disease, diabetes, hypertension, chronic obstructive pulmonary disease/asthma, autoimmune diseases history of cancer, immunocompromised, congestive heart failure, chronic kidney disease with dialysis, chronic kidney disease without dialysis, end-stage renal disease with dialysisMortalityMayer et al. (PMID: 33496668)Spain23,84442.3%49.93±19.4Retrospective studyOR: 1.19 (1.03–1.39)Age, sexDeathMuhammad et al. (PMID: 33538998)USA20060.5%58.9 ± 15.1Retrospective studyOR: 2.12 (0.94–4.77)Age, hypertension, coronary artery disease, chronic kidney disease, history of stroke, oxygen saturation, creatinine, blood urea nitrogen, creatine phosphokinase, troponin, procalcitonin, lactic acid, lactate dehydrogenase, C-reactive protein, initial d-dimer, ferritin, highest d-dimerMortalityYoshida et al. (PMID: 33546750)USA77647.3%60.5 ± 16.1Retrospective studyOR: 0.95 (0.53–1.71)Age, sex, hospital site, the charlson comorbidity indexDeathGirardin et al. (PMID: 33550849)USA444658.1%62±18Retrospective studyHR: 0.92 (0.79–1.06)Age, ethnic minority, male sex, low income, smoking, obesity, chronic obstructive pulmonary disease, asthma, sleep apnea, hypertension, diabetes, peripheral artery disease, coronary artery disease, autoimmune disease, cancerMortalityWargny et al. (PMID: 33599800)France279663.7%67.9 ± 13.2Retrospective studyOR: 1.15 (0.95–1.40)AgeDeathNote:§ The values are presented as mean ± standard deviation or median (interquartile range, IQR); USA, the United States of America; CI, confidence interval; OR, odds ratio; RR, risk ratio; RH, relative hazard; HR, hazard ratio. Open table in a new tab Note: We utilized Stata (version 12.1) for all statistical analyses. The pooled effect estimate and its 95% confidence interval (CI) were computed using a random-effects model. Inter-study heterogeneity was investigated using the cochrane Q test and I2 statistic, P < 0.1 or I2 > 50% shows a statistically significant heterogeneity. The statistical stability of the overall effects was assessed using leave-one-out sensitivity analysis. The risk of publication bias was evaluated using Begg's test. Subgroup analyses were carried out by sample size, age, male percentage, study design and effect estimate. Two-tailed P-value < 0.05 was considered statistically significant. Initial search yielded 2608 articles. After screening eligible articles according to inclusion and exclusion criteria, a total of twenty-seven studies composing of 146,364 cases were enrolled into this quantitative meta-analysis. Among the included studies, twenty-four studies were retrospective, one was prospective, one was longitudinal cohort study and one was nationwide cohort study. The sample sizes across the eligible studies ranged from 98 to 35,302. There were sixteen odds ratio (OR)-reported studies, nine hazard ratio (HR)-reported studies, one risk ratio (RR)-reported study and one relative hazard (RH)-reported study. The results of our pooled analysis are presented in Fig. 1A, which indicates that there was no significant relationship between dyslipidemia and COVID-19 morality (pooled effect size = 1.05, 95% CI [0.99–1.12], P = 0.100; I2 = 52.6%, random-effects model). Sensitivity analysis by deleting each study one by one demonstrated that our results were stable (Fig. 1B). When we limited dyslipidemia to hyperlipidemia, there was no significant relationship between hyperlipidemia and COVID-19 mortality (pooled effect size = 1.03, 95% CI [0.95–1.12]). We still observed no significant relationship between dyslipidemia and COVID-19 mortality in the subgroup analyses by age (pooled effect size = 1.08, 95% CI [0.99–1.18] for < 65 years old and pooled effect size = 1.02, 95% CI [0.93–1.12] for ≥ 65 years old), male percentage (pooled effect size = 1.04, 95% CI [0.96–1.13] for < 55% and pooled effect size = 1.08, 95% CI [0.97–1.20] for ≥ 55%), study design (pooled effect size = 1.06, 95% CI [0.99–1.14] for retrospective study), sample size (pooled effect size = 1.09, 95% CI [0.96–1.24] for < 1500 cases and pooled effect size = 1.04, 95% CI [0.96–1.14] for ≥ 1500 cases), and effect estimates (OR = 1.08, 95% CI [0.98–1.20] and HR = 1.02, 95% CI [0.92–1.13]). Begg's test indicated that there was no obvious publication bias (P = 0.505). This meta-analysis has several limitations that need to be mentioned:1Aung A.K. Aitken T. Teh B.M. Yu C. Ofori-Asenso R. Chin K.L. et al.Angiotensin converting enzyme genotypes and mortality from COVID-19: an ecological study.J. Infect. 2020; 81 (PubMed PMID: 33197472. Pubmed Central PMCID: 7666537): 961-965Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar most of the included studies were from USA, which limits its wider applicability of the present findings;2Grasselli G. Greco M. Zanella A. Albano G. Antonelli M. Bellani G. et al.Risk factors associated with mortality among patients with COVID-19 in intensive care units in Lombardy, Italy.JAMA Intern. Med. 2020; 180 (PubMed PMID: 32667669. Pubmed Central PMCID: 7364371): 1345-1355Crossref PubMed Scopus (786) Google Scholar the majority of studies were retrospective, thus further well-designed studies with more prospective researches are required to verify our results;3Mayer M.A. Vidal-Alaball J. Puigdellivol-Sanchez A. Marin Gomez F.X. Leis A. Mendioroz Pena J Clinical characterization of patients with COVID-19 in primary care in catalonia: retrospective observational study.JMIR Public Health Surveill. 2021; 7 (PubMed PMID: 33496668. Pubmed Central PMCID: 7871981): e25452Crossref PubMed Scopus (8) Google Scholar although the pooled effect estimate was calculated on the basis of adjusted effects, the adjusted variables are not completely consistent across the included studies;4Rosenthal N. Cao Z. Gundrum J. Sianis J. Safo S. Risk factors associated with in-hospital mortality in a US national sample of patients with COVID-19.JAMA Netw. Open. 2020; 3 (PubMed PMID: 33301018. Pubmed Central PMCID: 7729428)e2029058Crossref PubMed Scopus (229) Google Scholar only one included study explicitly states the specific type of dyslipidemia as total cholesterol, additional studies does not explicitly states the specific type of dyslipidemia such as abnormal levels of low-density lipoprotein, high-density lipoprotein, triglycerides and total cholesterol. Further studies should focus on the relationship between specific type of dyslipidemia and COVID-19 mortality when more data are available;5An C. Lim H. Kim D.W. Chang J.H. Choi Y.J. Kim S.W. Machine learning prediction for mortality of patients diagnosed with COVID-19: a nationwide Korean cohort study.Sci. Rep. 2020; 10 (PubMed PMID: 33127965. Pubmed Central PMCID: 7599238): 18716Crossref PubMed Scopus (91) Google Scholar the detailed information on medications for patients with pre-existing dyslipidemia is not available presently, thus we could not address the effects of medications on the relationship between dyslipidemia and COVID-19 mortality. In conclusion, our current study based on adjusted effect sizes demonstrated that dyslipidemia was not significantly associated with COVID-19 mortality. Further well-designed studies with large sample sizes are warranted to confirm our findings. The authors declare that they have no any potential conflict of interest regarding this submitted manuscript. We would like to thank Li Shi, Ying Wang, Jian Wu, Peihua Zhang, Yang Li and Wenwei Xiao (All are from Department of Epidemiology, School of Public Health, Zhengzhou University) for their kind help in searching articles and collecting data, and valuable suggestions for data analysis. Haiyan Yang and Yadong Wang designed the study. Hongjie Hou and Jie Xu performed literature search. Hongjie Hou and Haiyan Yang performed data extraction. Xuan Liang, Haiyan Yang, Hongjie Hou and Jie Xu performed statistical analyses. Haiyan Yang, Hongjie Hou and Yadong Wang wrote and reviewed the manuscript. All the authors approved the final version of the manuscript.