Association of cardiovascular disease and 10 other pre-existing comorbidities with COVID-19 mortality: A systematic review and meta-analysis

Background Estimating the risk of pre-existing comorbidities on coronavirus disease 2019 (COVID-19) mortality may promote the importance of targeting populations at risk to improve survival. This systematic review and meta-analysis aimed to estimate the association of pre-existing comorbidities with COVID-19 mortality. Methods We searched MEDLINE, SCOPUS, OVID, and Cochrane Library databases, and medrxiv.org from December 1st, 2019, to July 9th, 2020. The outcome of interest was the risk of COVID-19 mortality in patients with and without pre-existing comorbidities. We analyzed 11 comorbidities: cardiovascular diseases, hypertension, diabetes, congestive heart failure, cerebrovascular disease, chronic kidney disease, chronic liver disease, cancer, chronic obstructive pulmonary disease, asthma, and HIV/AIDS. Two reviewers independently extracted data and assessed the risk of bias. All analyses were performed using random-effects models and heterogeneity was quantified. Results Eleven pre-existing comorbidities from 25 studies were included in the meta-analysis (n = 65, 484 patients with COVID-19; mean age; 61 years; 57% male). Overall, the between-study heterogeneity was medium, and studies had low publication bias and high quality. Cardiovascular disease (risk ratio (RR) 2.25, 95% CI = 1.60–3.17, number of studies (n) = 14), hypertension (1.82 [1.43 to 2.32], n = 13), diabetes (1.48 [1.02 to 2.15], n = 16), congestive heart failure (2.03 [1.28 to 3.21], n = 3), chronic kidney disease (3.25 [1.13 to 9.28)], n = 9) and cancer (1.47 [1.01 to 2.14), n = 10) were associated with a significantly greater risk of mortality from COVID-19. Conclusions Patients with COVID-19 with cardiovascular disease, hypertension, diabetes, congestive heart failure, chronic kidney disease and cancer have a greater risk of mortality compared to patients with COVID-19 without these comorbidities. Tailored infection prevention and treatment strategies targeting this high-risk population might improve survival.


Introduction
The number of total cases of the coronavirus-2019 disease  continues to rise quickly, threatening thousands to millions of individuals with pre-existing chronic conditions who are disproportionately affected [1]. To date, as of July 9 th , 2020, the Johns Hopkins University coronavirus resource center reported that worldwide more than 180 countries have been affected with COVID-19 with more than twelve million confirmed cases and more than 500,000 deaths.
[2] As research related to potential risk factors for COVID-19 mortality continues, it is becoming clear that individuals with underlying comorbidities, such as cardiovascular diseases, hypertension, diabetes, congestive heart failure, cerebrovascular disease, chronic kidney disease, chronic liver disease, cancer, chronic obstructive pulmonary disease, asthma, and HIV/AIDS, may have a greater risk of death from COVID-19. [3,4] As the number of published studies increases, there is a widening gap due to inconsistent findings related to the influence of pre-existing comorbidities on COVID-19 mortality. Some studies report an association between pre-existing conditions and COVID-19 mortality, whereas others report no association. The differences may stem from the limited number of studies analyzed, heterogeneous methodologies and unaddressed sources of bias. Nevertheless, it is clear that regions experiencing the highest mortality rates, such as the United States, Europe, and China, also have the greatest burden of these pre-existing chronic conditions. [5] The novel virus, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of COVID-19, interacts with angiotensin-converting enzyme 2 (ACE2), a cellular binding site expressed in the heart, kidney and pulmonary alveolar type II cells. [6] It has been postulated, though not confirmed, that pre-existing use of angiotensin II type 1 receptor blockers (ARBs) may upregulate membrane-bound ACE2 hence increasing susceptibility to virus entry in humans. [7] Therefore, it is plausible that individuals with pre-existing chronic conditions such as hypertension and chronic heart failure taking ARBs may be more susceptible to the severity of SARS-CoV-2, including mortality.
To date, studies that have systematically explored the association of a range of pre-existing chronic conditions and COVID-19 mortality have limitations in the number of countries included, the number of studies included, and the number of conditions explored. [8,9]. Furthermore, these studies have significant unaddressed sources of bias that limit conclusions drawn from them. We took a comprehensive approach and estimated the association of major pre-existing chronic conditions, including cardiovascular diseases, hypertension, diabetes, congestive heart failure, cerebrovascular disease, chronic kidney disease, chronic liver disease, cancer, chronic obstructive pulmonary disease, asthma, and HIV/AIDS, and the risk of mortality from COVID-19. Although the majority of studies occurred in China, we identified additional studies involving patients from Europe, North America, and Africa.

Information source and search strategy
The present study has been registered with PROSPERO (registration ID: CRD42020187972). This study is being reported in accordance with the reporting guidance provided in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement and Meta-analysis of Observational Studies in Epidemiology (MOOSE). [10,11] We searched PubMed (MEDLINE), OVID (MEDLINE, HEALTHSTAR), SCOPUS, Joana Briggs International EBP, and Cochrane Library databases (from December 1 st , 2019, and July 7 th , 2020). We searched the grey or difficult to locate literature, including Google Scholar and Medrxiv. We performed hand-searching of the reference lists of included studies, relevant reviews, or other relevant documents. Studies reporting the risk of mortality in patients with COVID-19 were included. No limitations were applied on study design, country of publication, or language. Non-English-language articles were translated using language translation services at Penn State University Library and further included in the screening. Predefined search terms determined by the Medical Subject Headings (MeSH) and keywords included multiple combinations of the following: ("COVID-19" OR "coronavirus" OR "SARS-CoV-2" OR "2019-nCoV" OR "SARS nCoV2",) AND ("mortality") AND ("cardiovascular disease" OR "chronic obstructive pulmonary disease" OR "asthma" OR "coronary heart disease" OR "coronary artery disease" OR "hypertension" OR "diabetes" OR "congestive heart failure" OR "malignancy" OR "cancer" OR "chronic kidney disease" OR "chronic liver disease" OR "cerebrovascular disorders" OR "stroke "comorbidities" OR "asthma", OR" "HIV/AIDS" OR" "human immunodeficiency virus infection AND acquired immune deficiency syndrome"). Two reviewers (ESH and AES) independently screened titles and abstracts of the studies for inclusion eligibility. The comprehensive list of studies found as a result of our initial search was transferred into Endnote, which further removed duplicate studies.

Eligibility criteria
Studies were selected according to the following criteria: participants, exposure, comparison, condition or outcome(s) of interest, study design and context.
(ORs), risk ratios (RRs), or hazard ratios (HRs) or the data was presented such that the OR, RR, and HR could be calculated; the 95% CI was reported, or the data were presented such that the 95% CI could be calculated. We excluded case reports or case series with less than 10 patients, studies not conducted on humans, review papers, meta-analyses, literature reviews, and commentaries. Excluded studies were documented with reasons for their exclusion.

Data extraction
Two reviewers (ESH and AES) initially screened titles and abstracts of all identified articles for eligibility. After initially screening articles for inclusion based on titles and abstracts, full-text articles were screened. Disagreements were resolved by discussion to meet a consensus. If necessary, a third reviewer (PS) was consulted in order to reach a consensus. We extracted the following information: year of publication, date of the study, the sample size of patients with COVID-19, number of participants with each comorbidity who died and did not die, the point estimates and 95% confidence interval of mortality of patients with COVID-19 with each underlying condition (S1 Table), the mean or median age with their corresponding standard deviation or interquartile range, respectively, the proportion that was male, and the covariates adjusted for each study. We gave priority to adjusted estimates if available.

Study quality assessment
Two reviewers (EH and AES) independently assessed the quality of the included studies. The Newcastle-Ottawa Scale (NOS) was utilized for the quality assessment of the included studies. [13] NOS scale rates observational studies based on 3 parameters: selection, comparability between the exposed and unexposed groups, and exposure/outcome assessment. It assigns a maximum of 4 stars for selection, 2 stars for comparability, and 3 stars for exposure/outcome assessment. Studies with less than 5 stars were considered low quality, 5-7 stars of moderate quality, and more than 7 stars of high quality. In reporting results, we gave preference for results from clinical trials over observational studies.

Data synthesis and statistical analysis
We adopted a narrative approach to describing the number of studies, study settings, proportion of sex, mean or median age and covariates adjusted for in each study. Our primary outcome was the risk of mortality in patients with COVID-19 associated with pre-existing chronic diseases. We combined RRs and HRs with ORs in the present meta-analysis and reported the pooled effect size as RRs as common risk estimates for all studies.
For studies without measures of associations, a generalized linear mixed model was used to calculate the OR using the number of events and the sample size of each study group. [14]. Effect sizes were log-transformed to normalize the distributions. Next, standard errors (SEs) were calculated via the following equations [15]: Lower = log (lower 95% CI) and upper = log (upper 95% CI), and SE = (upper-lower)/3.92. To assess the associations between pre-existing conditions and the risk of mortality, we pooled the RR estimates for the presence versus absence of pre-existing conditions from each study, weighted by the inverse of their variances (inter-study plus intra-study variances). The metagen function from the R package meta was used to calculate the pooled effect estimates using random-effects models. [16] The DerSimonian and Laird (DL) random-effects method was used to estimate the pooled inter-study variance (heterogeneity). [17] We graphically displayed individual and pooled estimates with forest plots. Inter-study heterogeneity was assessed using I 2 statistics, expressed as % (low (25%), moderate (50%), and high (75%) and Cochrane's Q statistic (significance level < 0.05) [18,19]. For outcomes with high between-study heterogeneity, influence sensitivity analysis was undertaken to explore the effect of each individual study on the overall pooled estimate. [20] If a specific comorbidity had 10 or more studies and significant between-study heterogeneity was observed, we conducted meta-regression analysis to explore sources of variation, using mean age and proportion of males as regressors. [16] Potential ascertainment bias (as might be caused by publication bias) was assessed with funnel plots, by plotting the study effect size against standard errors of the effect size, and Egger's test. [21] Trim and fill analyses using Duval and Tweedie non-parameteric method were used to adjust for the publication bias. [22] All statistical analyses were performed with R software, version 3.4.3 (R, College Station, TX).

The risk of mortality in patients with COVID-19 with cardiovascular disease
Of the fourteen studies that reported the risk of mortality associated with pre-existing cardiovascular disease, 7 reported a statistically significant association. The RR point estimates for mortality from COVID-19 and cardiovascular disease ranged from 0.72 to 8.89 (Fig 2). The overall pooled RR of COVID-19 mortality associated with cardiovascular disease was 2.25 (95% CI: 1.60-3.17), implying an approximately 2-fold greater risk of death. Between-study variation was moderate (I 2 = 49, p = 0.02).  (Fig 3). Cerebrovascular disease demonstrated a higher risk of death, though it was not a statistically significant finding (2.16 [95% CI 0.97 to 4.80]). In addition to this, chronic liver disease, COPD, asthma, and HIV/AIDS comorbidities were not significantly associated with greater risk of mortality.

Publication bias, study heterogeneity, and meta-regression
The risk of publication bias was found to be significant for hypertension, cerebrovascular disease, and cancer (Table 2; the funnel plots are provided in S1 Fig). Trim and fill methods were used to adjust for the publication bias [22]. If the asymmetry is due to publication bias, our analyses suggest that the adjusted effect estimates for hypertension remained significant, However, the adjusted effect estimates for cancer were no longer significant.
Significant between-study heterogeneity was observed for cardiovascular disease, hypertension, diabetes, cerebrovascular disease, chronic kidney diseases, COPD, and HIV/AIDS. If a

PLOS ONE
specific comorbidity had more than 10 studies and exhibited significant medium or high between-study heterogeneity, we further conducted meta-regression to explore the sources of heterogeneity. [46] Mean or median study age and the proportion of males were regressors in the univariate meta-regression. Results from the meta-regression analyses showed no statistically significant association between either mean study age or the proportion of males with estimated RRs for mortality (Table 3). We further performed influence sensitivity analyses for outcomes with high between-study heterogeneity by excluding and replacing one study at a time (Leave-One-Out method) from the meta-analysis and calculated the RR for the remaining studies [20]. No substantial change from any of the pooled RR were observed when other studies were removed in turn, indicating that no individual study had a considerable influence on the pooled estimate. The plots for the analysis estimates are provided in S2 Fig.

Principal findings
The results of this systematic review and meta-analysis suggest that hospitalized patients with COVID-19 with pre-existing cardiovascular disease, hypertension, diabetes, congestive heart failure, chronic kidney disease and cancer have a greater risk of death from COVID-19.

Strengths and limitations
This review provides up-to-date results of comorbidities' influence on the higher risk of mortality in patients with COVID-19 by synthesizing a large number of recently published studies. The study yielded a large number of individuals from countries representing Europe, North

PLOS ONE
America, Africa, and Asia. Nevertheless, this meta-analysis had several limitations. While some studies reported the race and ethnicity of study participants, others provided little or no information. Thus, we could not investigate race and ethnicity's influence, which could have been a cause of variation between studies. Due to a lack of individual patient-level data, we used study level data in meta-regression analyses to assess heterogeneity between studies. We were unable to identify the main sources of heterogeneity in the effect estimates. Perhaps future analysis could use individual patient-level data to explore sources of heterogeneity.
Most of the studies included were at low risk of bias. Nevertheless, the NOS used to assess the included studies' quality has low or unknown validity. [49] The alternative validated scale with high quality, the Risk of Bias in Nonrandomized Studies of Interventions (ROBINS-I), [50] could not be used in our meta-analysis because included studies did not involve any interventions. Nevertheless, we carefully evaluated included studies and listed potential confounders adjusted for each study that reported adjusted estimates. Furthermore, we investigated and reported adjusted effect estimates for the outcomes that showed potential publication bias (as quantified by a significant Egger's test and asymmetrical funnel plots). Many of the included studies reported effect size estimates from all covariates within each model, which can induce multiplicity and lead to "the Table 2 fallacy". [51] From a causal inference perspective, this is a significant source of bias and could affect the interpretation of the results in causal terms. Many studies reporting COVID-19 mortality have been retrospective cohort studies or case-control studies that report odds ratios from logistic regression models, and yet COVID-19 mortality is relatively common (>10%). Since odds ratios of fairly common outcomes have the potential for biasing effect estimates towards a stronger association, the magnitude of the pooled estimates from our meta-analysis could be inflated. [52] Comparison with other studies Our study findings are similar to results from previously published systematic reviews, suggesting a higher mortality rate of COVID-19 individuals with cardiovascular, chronic kidney disease and cancer. The overall 95% confidence interval obtained by Singh and colleagues overlapped with the one we identified. Pranata and colleagues assessed studies published until April 10 th , 2020, and included 4448 patients from 16 studies. [53] They found a 2-fold greater risk in mortality in patients with COVID-19 with cardiovascular and cerebrovascular disease. In the current meta-analysis, the risk of mortality was approximately 2-fold higher in individuals with cerebrovascular disease. However, the association was not statistically significant. In addition to what is reported in published studies, this systematic review and meta-analysis incorporated evidence from the most recent studies, a large sample size, and a significant and higher estimated mortality risk of hypertension, diabetes, congestive heart failure, and chronic kidney disease than found previously.
A probable hypothesis for the pathophysiological mechanism related to the higher risk of mortality among patients with COVID-19 with pre-existing chronic conditions may stem from the increased allostatic load. Chronic conditions cause dysregulation of major physiological system, including the hypothalamic-pituitary-adrenal axis, the sympathetic nervous system, and the immune system. [54] The chronic nature of such conditions induces the ''wear and tear" on the body's regulatory system, [55] leading to the accumulation of pro-inflammatory cytokines, which affects the cellular immune system. As a result of the reduced immunity, these individuals become very susceptible to severe complications of SARS-CoV-2 and death. Such an association with this type of virus is not relatively new. Seasonal influenza, SARS-CoV, and Middle Eastern Respiratory Syndrome-CoV are also associated with increased severity and mortality in patients with preexisting conditions. [29,56]

Implications for research and clinical practice
As the race towards developing a vaccine against SARS-CoV-2 intensifies, our results support the notion that individuals who appear to be at increased risk should receive immunization priority. Individuals with pre-existing cardiovascular disease, hypertension, diabetes, congestive heart failure, chronic kidney disease and cancer are at heightened risk of death from the virus, thus should be given a priority during the allocation of a vaccine, especially if it is in limited supply. Historically, targeted public health vaccination intervention strategy for influenza vaccination is recommended by the Advisory Committee on Immunization Practices against seasonal influenza. [57] In the population with chronic comorbidities, annual influenza vaccination significantly reduces mortality and morbidity. [58] Mounting evidence postulates that SARS-CoV-2 may become seasonal requiring annual vaccination. [59] The majority of patients with pre-existing cardiovascular disease, hypertension, diabetes, chronic kidney disease and congested heart failure use renin-angiotensin-aldosterone system (RAAS) blockers, which are postulated to increase the risk of developing a severe and fatal SARS-CoV-2 infection. [60] In experimental studies by Ferrario and colleagues, ACE inhibitors or ARBs increased cardiac levels of Ace2 mRNA compared with a placebo. [61] Notably, cardiac levels of Ace2 mRNA increased by 4.7-fold or 2.8-fold with either lisinopril (an ACE inhibitor) or losartan (an ARB), respectively. Nonetheless, recent clinical studies have found no association between the RAAS blockers and increased mortality from COVID-19. [62,63] Such clinical or experimental evidence suggests that RAAS is an appropriate drug to manage hypertension and other cardiovascular diseases in the setting of COVID-19.

Conclusion
Our findings suggest that of the major comorbidities analyzed, cardiovascular disease, hypertension, diabetes, congestive heart failure, chronic kidney disease, and cancer carry the highest risk of death from COVID-19. This research highlights the importance of paying attention to the populations with these specific comorbidities and tailor treatment to meet their need.