Anxiety and prognosis of patients with myocardial infarction: A meta‐analysis

Abstract Although anxiety is highly prevalent after myocardial infarction (MI), but the association between anxiety and MI is not well established. This study aimed to provide an updated and comprehensive evaluation of the association between anxiety and short‐term and long‐term prognoses in patients with MI. Anxiety is associated with poor short‐term and long‐term prognoses in patients with MI. We performed a systematic search in the PubMed and Cochrane databases (January 2000–October 2020). The study endpoints were complications, all‐cause mortality, cardiac mortality, and/or major adverse cardiac events (MACEs). Pooled data were synthesized using Stata SE12.0 and expressed as risk ratios (RRs) and 95% confidence intervals (CIs). We included 9373 patients with MI from 16 published studies. Pooled analyses indicated a correlation between high anxiety and poor clinical outcomes (RR: 1.19, 95% CI: 1.13–1.26, p < .001), poor short‐term complications (RR: 1.23, 95% CI: 1.09–1.38, p = .001), and poor long‐term prognosis (RR: 1.27, 95% CI: 1.13–1.44, p < .001). Anxiety was also specifically associated with long‐term mortality (RR: 1.16, 95% CI: 1.01–1.33, p = .033) and long‐term MACEs (RR: 1.54, 95% CI: 1.26–1.90, p < .001). This study provided strong evidence that increased anxiety was associated with poor prognosis in patients with MI. Further analysis revealed that MI patients with anxiety had a 23% increased risk of short‐term complications and a 27% increased risk of adverse long‐term prognosis compared to those without anxiety.


| Aim
This study was a meta-analysis of prospective studies. We sought to compare the clinical outcomes between MI patients with and without anxiety. We included all studies available from January 2000 to October 2020. Systematic identification, appraisal, synthesis, statistical aggregation, and reporting of results were performed in accordance with known guidelines. 30,31 2.2 | Literature search A systematic search was performed in the PubMed and Cochrane databases using keyword-based queries with the following terms: "myocardial infarction" and "anxiety." The results were limited to human studies published in English. Moreover, reference lists of original and review articles were manually screened as a supplement. This meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. 32

| Selection
Studies were identified by two independent raters, and any disagreement between them was resolved via consensus. Articles meeting the following criteria were included: (1) prospective studies that enrolled patients with established AMI; (2) AMI patients diagnosed with anxiety using reliable and validated instruments; (3) studies including clinical outcomes in patients and documenting the risk ratios (RRs) and their corresponding 95% confidence intervals (CIs) for clinical outcomes in patients with anxiety compared with those in patients without anxiety.

| End points
The prognosis of this meta-analysis focused on the short-term complications and long-term mortality or adverse events. The study endpoints were complications, all-cause mortality, cardiac mortality, and/or major adverse cardiac events (MACEs). Complications include recurrent ischemia, or reinfarction, or ventricular arrhythmia, or congestive heart failure (CHF), or death during hospitalization. MACEs include cardiac death, reinfarction, readmission for recurrent ischemia, and ventricular arrhythmia. Recurrent ischemia was defiend as new onset of chest pain, ST-segment changes on the electrocardiogram, or hemodynamic instability. Reinfarction was evidenced by recurrent positive creatine kinase-MB. Ventricular arrhythmia was defined as ventricular tachycardia lasting more than 15 s, or any ventricular tachycardia requiring pharmacological and/or electrical intervention due to hemodynamic instability or chest pain, or ventricular fibrillation. CHF was identified either through a diagnosis in the discharge summary, or both (a) chest X-ray consistent with CHF and (b) acute administration of diuretic or other agent to treat CHF.

| Statistical analysis
The pooled RR and its corresponding 95% CI for prognosis were calculated using Stata SE12.0 (StataCorp., College Station, Texas). The I 2 test and Q statistic test were used to assess the heterogeneity among the included studies. When there was no obvious heterogeneity, pooled data were analyzed using a fixed-effects model; otherwise, the random-effects model was used (p < .1 and/or I 2 > 50% suggested significant heterogeneity). Publication bias was evaluated by visual inspection of the funnel plots. Sensitivity and subgroup analyses were performed to evaluate the stability of the results. The "adjusting for depression" was defined as depression was included in the adjustment factors of including studies. A p-value of <.05 was considered statistically significant.

| RESULTS
A flowchart of the search results is shown in Figure 1. A total of 2302 studies were initially searched. After screening the title, abstract, and full text, 16 studies were finally included in this meta-analysis. 14-29

| Quality assessment
The methodological quality of eligible studies was evaluated using the Newcastle-Ottawa quality assessment scale (NOS). 33 The methodological quality of the studies were judged using three main criteria: the selection of the study groups, comparability of the groups, and ascertainment of either the exposure or outcome of interest for studies.
The total NOS score was ranged from 1 to 9. Studies with NOS scores ≥6 were classified as high-or medium-quality studies, while studies with NOS scores <6 were classified as low-quality studies. The quality assessment of the 16 included studies is shown in Suppl Table 1.

| Anxiety and clinical outcomes in patients with MI
The 16 included studies reported a combined end-point. We chose cardiac mortality as the end-point if the study reported more than one end-point. There was significant heterogeneity among the studies (I 2 = 40.3%, P h = 0.048); the pooled RRs for complications and allcause mortality and cardiac mortality and MACEs with corresponding 95% CIs were evaluated using a random-effects model. The overall results demonstrated that MI patients with anxiety had a significantly poorer prognosis than those without anxiety (RR: 1.27, 95% CI: 1.15-1.40, p < .001; Figure 2(A)). Further analysis of excluding studies in which anxiety were evaluated before MI showed a consistent result (RR: 1.26, 95% CI: 1.13-1.41, p < .001; Figure 2(B)). The RRs for sensitivity analysis ranged from 1.25 to 1.31, which showed that the combined result was not significantly altered after excluding any studies. Subgroup analysis including 13 studies with multivariable RR also showed that anxiety was associated with poor prognosis in patients with MI (RR: 1.27, 95% CI: 1.14-1.41, P < 0.001; Figure 2

| Anxiety and long-term prognosis in patients with MI
Twelve studies, comprising 8204 patients with MI, reported the longterm prognosis for all-cause and cardiac mortality and MACEs. We

| Anxiety and long-term mortality in patients with MI
Seven studies, including 6760 patients with MI, provided information on long-term mortality analysis, which included all-cause mortality and cardiac mortality. If the research reported both all-cause mortality and cardiac mortality as endpoints, we chose cardiac mortality. The pooled results of the fixed-effects model (I 2 = 0.0%, P h = 0.501) showed a significant association between anxiety and long-term mortality (RR: 1.16, 95% CI: 1.01-1.33, p = .033; Suppl Figure 1A). The RRs for sensitivity analysis ranged from 1.08 to 1. 19

| Anxiety and long-term cardiac mortality in patients with MI
Five studies comprising 3688 patients with MI reported the association between anxiety and long-term cardiac mortality. Since no significantly heterogeneity was observed among studies (I 2 = 0.0%, P h = 0.985), a fixed-effects model was used. The pooled results showed that there was no significant association between anxiety and long-term cardiac mortality (RR: 1.12, 95% CI: 0.95-1.32, p = .165; Suppl Figure 3). The RRs for sensitivity analysis ranged from 1.08 to 1.13, which showed that the combined result was not significantly altered after excluding any studies.

| Anxiety and long-term MACEs in patients with MI
We evaluated the correlation between anxiety and long-term

| Publication bias
For the meta-analysis of the association between anxiety and combined clinical outcomes and long-term prognosis and mortality, the funnel plots were symmetrical and are provided in Figure 4  respectively. Furthermore, we found that anxiety was associated with a 1.14-times higher long-term all-cause mortality rate and 1.12-times higher long-term cardiac mortality rate than those without anxiety.
The number of studies (n = 16) included in our meta-analysis was greater than that in the abovementioned two previous meta-analyses.
There was no significant heterogeneity with I 2 < 50%, except shortterm prognosis and long-term all-cause mortality.  14,16,23 Some studies found that the association between anxiety and long-term prognosis was insignificant after adjusted for depression. 23,34 And other studies shown the different results. 14,16 So in order to explore the potential influence of depression between anxiety and prognosis in patients with MI. We performed subgroup analyses after adjusting for depression; we observed no significant association between anxiety and poor clinical outcomes, poor long-term prognosis, poor long-term mortality (p > .05). However, there was a significant association between anxiety and poor long-term MACEs (p = .002). The results suggest that depression may be a contributing factor for the association between anxiety and prognosis in patients with MI. However, we included only five studies in the subgroup analysis; hence, the results may be unreliable, and further analyses should be performed with a larger sample size.
In two previous meta-analyses, MI patients with anxiety had a poorer prognosis than those without anxiety. 34,35 In one meta analysis, patients with MI and unstable angina were included, the results were assessed using bivariate analysis, and there was a significant publication bias in the combined outcomes. 35 The other metaanalysis evaluated the effect of anxiety in patients with ACS, and included patients with depression or other factors. 34 In addition, both the meta-analyses focused on only long-term prognosis, and did not analyze the association between anxiety and the short-term prognosis of patients with MI. Therefore, we performed an updated metaanalysis to compare the clinical outcomes in MI patients with and without anxiety. Compared to the aforementioned two previous meta-analyses, our updated meta-analysis has three strengths. First, we only included MI patients, with no other pathology like unstable angina. Second, we only considered anxiety as the exposure factor, and studies that considered other exposure factors, such as depression or perceived control, were excluded. Third, in addition to longterm prognosis, we also paid attention to short-term prognosis.
Furthermore, various clinical outcomes were evaluated in MI patients with and without anxiety. Overall, our meta-analysis provided more reliable results than the aforementioned studies and reinforced previous findings that showed that anxiety is associated with a mild risk for poor prognosis.
The pathophysiological mechanism underlying the adverse association between anxiety and cardiac prognosis is complicated and unclear. 36 Some hypotheses, such as neuro-humoral mechanisms, platelet activation, endothelial dysfunction, and multi-organ network, may help explain this association. Because of decreased vagal tone, anxiety is related to reduced baroreflex control of the heart and increased arrhythmia in patients with AMI. 37 Anxiety is also potentially related to hypothalamic-pituitary-adrenal axis dysregulation. 38 Platelet aggregation is a main contributor to the development of MI. Anxiety is associated with higher platelet reactivity, thus enhancing platelet aggregation and leading to clot formation. 39,40 Anxiety is associated with poor endothelial and vascular smooth muscle function in patients with atherosclerosis and has been shown to increase vascular events in patients with MI. 41 Recently study have highlighted the multi-organ network linking brain emotional neural activity, macrophage hematopoiesis, and arterial inflammation of AMI. Brain emotional neural activity is closely linked with acute plaque instability through enhanced hematopoietic macrophage activity. 42 This updated meta-analysis has some limitations. First, the number of included studies was relatively small, and the further studies with a larger sample size is needed. Second, anxiety was evaluated using nine different instruments in the included studies, with no agreed standard to judge whether the patient had anxiety. The timing of evaluation was different in the 16 studies; two studies evaluated anxiety prior to MI based on patients' memories. 15,26 Third, the follow-up duration was variable in the included studies, ranging from 1.3 to 10 years. Fourth, publication bias may exist; however, no significant publication bias was observed based on the funnel plot and the sensitivity analysis showed stable results. Overall, all these limitations may have affected the results.
In conclusion, our study provided strong evidence that increased anxiety was associated with poor prognosis in patients with MI. Well-F I G U R E 4 Funnel plots of the meta-analysis of anxiety and prognosis in patients with MI. Clinical outcomes (A), long-term prognosis (B). MI, myocardial infarction designed clinical studies including a larger sample size and other ethnic groups are needed to further confirm the association between anxiety and prognosis in patients with MI, to determine the optimal management methods for MI patients with anxiety, which can improve their prognosis.

CONFLICT OF INTEREST
The author declare no potential conflict of interest.

DATA AVAILABILITY STATEMENT
The data supporting this meta-analysis are from previously reported studies and datasets, which have been cited.