Influenza vaccine and cardiac protection: a study from a tertiary care center

ABSTRACT

Patients with heart disease (HD) are at increased risk of developing cardiac complications if they acquire the influenza virus. The objective of this study was to determine whether the influenza vaccine has a primary role in preventing newly diagnosed HD in patients who have no history of HD and who were being followed up at the American University of Beirut Medical Center (AUBMC). The study is a retrospective cohort, with 2-years follow up, which was conducted using electronic medical records between the years of 2011–2013 in a tertiary care center. All patients 60 years and older (n = 698) who have taken the flu vaccine were randomly selected from the University Health Service records (UHS) and compared to a group who has not taken the flu vaccine during the same period. The odds of developing HD among vaccinated people with cofactors are 0.97 times the odds of that among non-vaccinated. This odds ratio is not significantly different than that of people vaccinated without cofactors (OR = 1.74). The occurrence of HD in the presence of vaccination revealed a non-significant decrease trend with the increase in a number of risk factors (OR = 1.61 vs 0.97). Our results suggest that there was a non-significant difference between the effect of the vaccine for influenza on patients who had cofactors for HD and those who had not. Similarly, the effect of vaccine showed a non-significant increase in protective effect as the number of comorbidities increase. The potential effects of the vaccine may be related to the protection against flu.

KEYWORDS:

• Influenza virus
• vaccine
• heart disease
• elderly
• risk factors
• prevention

Introduction

Influenza (flu) is a contagious respiratory illness in humans caused mainly by two types of RNA viruses,¹ influenza A and B, which can trigger epidemic human disease in a seasonal pattern.² The symptoms of influenza include cough, headache, myalgia, joint pain, congestion, sore throat, shortness of breath, nausea and other irritating symptoms. The illness can be mild and self-limiting or severe leading to serious outcomes such as hospitalization or even death.¹ In the absence of influenza vaccine conferring permanent immunity, influenza remains one of the prominent causes of prevalent viral diseases, with high morbidity and mortality. Globally, an estimation of 5–10% of adults and 20–30% of children are infected by influenza annually, whereas an estimated 250,000–500,000 severely infected persons die annually.³ Vulnerable populations to influenza infection include elderly, young children, and people with certain health conditions.¹ Elderly specifically, are at increased risk of co-morbid conditions, such as heart disease (HD), pulmonary diseases, immune-compromised state, and other chronic diseases.⁴ In such high-risk individuals, the influenza virus becomes more than just ”flu”, as it can lead to serious and life-threatening complications such as pneumonia, myocardial infarctions, and exacerbation of comorbid conditions.⁵ Seasonal peaks of all-cause and cardiovascular mortality overlap with seasonal peaks in influenza and pneumonia.^4,6,7 Several observational studies have proposed a causal relationship between influenza activity and ischemic heart disease. Gwini et al. reported that there was a decrease in the relative incidence of acute myocardial infarction (AMI) following vaccination for influenza.⁸ Clayton et al. showed that risks of myocardial infarction and stroke were increased within 7 days of respiratory infection.⁹ A recent study revealed that autopsy-confirmed coronary heart disease was increased in the 10 weeks after seasonal influenza activity.¹⁰ Evidence from the second wave of the influenza pandemic in 2009 that occurred outside the winter season, observed and detected the co-occurrence of a peak of cases of AMI.^{11 .}Though some studies have reported that influenza infection leads to an excess in specific cardiac mortality^10,11 the influence of the influenza vaccination in escaping mortality among patients with chronic heart diseases is not well recognized.¹² Despite the abundant studies on the effect of influenza vaccine^13–16 on cardiac disease, few studies looked at the effectiveness of the influenza vaccine in primary prevention of heart disease. Most of these studies suffered risk of bias and their evidence was not sufficient to establish that the influenza vaccine can be used as a tool for primary prevention of heart disease.^15,16 In this study, we aim to determine whether the influenza vaccine has a primary role in preventing newly diagnosed heart disease in patients who have no previous history of heart disease and who were followed up for 2 years after receiving the vaccine.

Materials and methods

The study was conducted at a tertiary academic center at the American University of Beirut University Health Services outpatient clinics. The study was a retrospective research that depended on chart review. All information was retrieved from the electronic medical records that included all patients’ visits to the outpatient clinic and admissions to the nearby affiliated hospitals between the years of 2011–2013. The study was approved by the Institutional Review Board of the institution.

Participants

Patients who are registered in Health Insurance Plan (HIP) aged ≥60 years with no previous history of heart disease and who have received the influenza vaccine during the years of 2011–2013 were selected and they constituted the vaccinated group. The non-vaccinated group constituted of randomly selected elderly patients of the same age group, who have not received the flu vaccine within the same period of time (2 years).

Inclusion criteria were patients who were 60 years and older and who are members of the HIP at the AUBMC. Exclusion criteria include anyone who has a previous history of heart disease and anyone who has received the influenza vaccine during 2011–2013. Our sample size was a total of 698 participants. Among those 698 patients, only 108 were vaccinated and the rest (590) were not vaccinated. Besides demographic information, data related to the presence of comorbidities i. e hypertension (HTN), diabetes (DM), dyslipidemia (DL), smoking habits, vaccination status, and hospital admissions due to cardiac disease or events in the year following vaccination were retrieved from the patients’ charts on the electronic medical record.

Heart disease: outcome variable

Cardiac disease was defined by ICD 9 codes and it included all types of myocardial infarctions, post-myocardial infarction syndrome, sub-acute forms of ischemic heart disease, old myocardial infarctions, and different forms of angina, coronary atherosclerosis/occlusion, and congestive heart failure. The main independent variable was whether the patient had received the influenza vaccine or not. Besides the main independent variable, other factors were also considered in the analyses. Gender, age, marital status, presence of hypertension, dyslipidemia, diabetes, and smoking were all co-variables adjusted for in our analyses. This information is present in the electronic record of the University Health Service records (UHS). Hypertension is defined as blood pressure (BP) ≥140 mmHg systolic and/or ≥90 diastolic and/or currently taking antihypertensive medications in National Health and Nutrition Examination Surveys (NHANES).¹⁷

Dyslipidemia is defined as elevations of fasting total cholesterol concentration which may or may not be associated with elevated triglyceride concentration.¹⁸

Statistical analyses

For analysis of data, SAS was used to compute the statistical analyses. When data were collected, information about 725 patients were retrieved. Data about the development of heart disease in the year after vaccination whether the patient has developed it or not, was not present for all of those 725, so every patient with missing data about heart disease was removed from the analyses. So the final sample size was 698 patients. We used simple logistic regression to study the effect of vaccine on heart disease and the effect of every cofactor, including HTN, DM, DL, smoking, age and gender, on heart disease. We also used multiple logistic regressions to evaluate the effect of vaccine on heart disease while adjusting for all of the previously mentioned cofactors. Logistic regressions were used given that our outcome variable (HD) was binary, noted as yes or no, in order to calculate the odds ratio of developing heart disease and to study the relationship or effect of the cofactors on developing HD. Assumptions of the statistical tests including having a dichotomous outcome variable, absence of outliers in the data as well as the absence of high correlations between the predictors were all met before analyses were conducted.

Results

Demographic characteristics of the sample

The demographic results of the study population are summarized in Table 1. The mean (SD) age of this sample of elderly was 68.5 (3.6) years. Of these subjects, 59% were female and 87.1% were married. 41.3% were hypertensive, 23% were diabetic, 29.5% were dyslipidemic and 29.9% were a smoker. 8.02% of our study population has developed heart disease.

Table 1. Descriptive characteristics of the study population.

		Totaln = 698	Not Vaccinatedn = 590	Vaccinatedn = 108
Age Mean (SD)		68.5 (3.6)	68 (2.8)	70.9 (5.8)
Sex %	Female	59	58	64.8
	Male	41	42	35.2
Marital%	Married	87.1	88.1	81.5
	Single	9.2	8.8	11.1
	Widowed	2.2	2.5	7.4
	Divorced	0.4	0.5
HTN%	N	58.7	61	46.3
	Y	41.3	40	53.7
DM %	N	76.9	78.6	67.6
	Y	23.1	21.4	32.4
DL %	N	70.5	72.5	59.3
	Y	29.5	27.5	40.7
HD %	N	92	92.7	88
	Y	8	7.3	12
Smoker %	N	70.1	70.3	68.5
	Y	29.9	29.7	31.5

Association between heart disease and vaccine

Table 2 shows the associations between heart disease and vaccine in terms of odds ratios and confidence intervals in two different models, the first (Model 1) containing all of the comorbidities and vaccine simultaneously, and the second (Model 2) studying the relationship of each of the explanatory variables on heart disease separately.

Table 2. Odds ratios (95% CI) for associations between heart disease and vaccine after adjusting for other covariates in model 1 and the effect of vaccine and every covariate independently on heart disease in model 2.

	Model 1		Model 2
	Heart Disease		Heart Disease
Independent variables	OR	95% CI	OR	95% CI
Vaccine	0.97	0.45–2.06	1.74	0.9–3.36
HTN	4.53	2.07–9.88*	7.603	3.76–15.34*
DL	1.33	0.71–2.49	3.055	1.75–5.31*
DM	2.05	1.08–3.86*	4.15	2.37–7.25*
Sex	0.12	0.04–0.38*	0.288	0.14–0.58*
Smoking	2.39	0.9–6.32	0.688	0.36–1.30
age	0.89	0.82–0.97*	0.96	0.90–1.02

*: Odds ratios that are statistically significant

In the second model, those who have received the vaccine were 74% more likely to develop HD than those who have not received the vaccine. This association was not statistically significant. The odds of developing HD among hypertensive, dyslipidemic and diabetic were higher than the odds of developing HD among non-hypertensive, non-dyslipidemic and non-diabetic.). All of these associations were statistically significant . The odds of developing HD among men were 0.28 times the odds of developing HD among women after adjusting for other covariates. This relationship was also statistically significant.

In the first model, after adjusting for all of the co-variables, the odds of developing HD among vaccinated people were 0.97 times the odds of developing HD among non-vaccinated. This odds ratio is not significantly different than that of when the vaccine was studied by itself in the first model (OR = 1.74). It is another way of saying the odds of developing HD among diabetics and hypertensive are 2.05 and 4.35 times the odds of developing HD among non-diabetics and non-hypertensive, respectively, after adjusting for other co-variables. These relationships were statistically significant. In addition, the odds of developing HD among male is 0.12 times the odds of developing HD among females after adjusting for other covariates. This association is also statistically significant. For every one-unit increase in age, there’s an 11% decrease in the risk of developing HD after adjusting for other covariates. This relationship was statistically significant.

Table 3 below shows the change in the effect of vaccine on heart disease after adding the risk factors, i.e. change of effect of vaccine on heart disease in the presence of one cofactor as compared to the effect in the presence of multiple cofactors. As shown in Table 3, the OR is decreasing with the addition of comorbidities to the model. For example, when one risk factor was included, the odds of developing HD among people who have this cofactor was 1.61 times the odds of developing HD among people who do not have it. This odds ratio kept on decreasing to 0.97 after adding all of our comorbidities to the model. None of these associations were statistically significant.

Table 3. The change in the effect of vaccine on heart disease after adding each cofactor at a time.

Heart Disease
	Odds ratio (OR)	95% CI
Vaccine + 1 risk factor
Vaccine	1.61	0.80–3.24
Vaccine + 2 risk factors
Vaccine	1.33	0.65–2.71
Vaccine + 3 risk factors
Vaccine	1.33	0.66–2.69
Vaccine + 4 risk factors
Vaccine	1.32	0.65–2.69
Vaccine + 5 risk factors
Vaccine	0.98	0.46–2.09
Vaccine + 6 risk factors
Vaccine	0.97	0.45–2.06

Discussion

In this study, we looked at the association between influenza vaccination and the occurrence of cardiovascular events in patients 60 years and older who have a negative history of heart disease and followed up at a tertiary care center. Our study sheds light on important aspects in the design of research when studying influenza vaccination among the adult population and the effect of the vaccine in primary prevention of cardiac disease. Cardiac complications from influenza infection are well documented and it is apparent from previous studies that there is a potential association between influenza infections and acute myocardial infarctions.^19–21 Because influenza often exacerbates underlying diseases, i.e., cardiac disease that may be recorded as the primary cause of death^5,9,10 preventing influenza infection becomes vital in reducing severe illness in patients with co-morbid conditions. This study showed clearly that each of the co-morbid conditions of hypertension, dyslipidemia, diabetes besides age increased significantly the odds of having heart disease. This is in complete agreement with well-established literature about the risks for coronary heart disease²² and this validates the accuracy of our data. While influenza vaccination may reduce cardiovascular mortality and combined cardiovascular events in patients with cardiovascular disease,²³ the evidence for the role of influenza vaccination in the primary prevention of cardiovascular disease is not available.²³ Irrespective of co-morbid conditions and when standardizing for risk factors, our data showed that patients who were vaccinated were significantly more likely to suffer heart illness than those who were not vaccinated. Several biases can be confounding our results. First, because there is no national policy to adopt influenza vaccination in adults, treating physicians are left to their discretion to decide whom to vaccinate. This indication bias results in physicians offering vaccination for influenza to those who were considered to be high risk leaving those with less comorbid conditions with no vaccination. Second, we have the “frailty bias” that can underline the fact that worsening of health is associated with increasing vaccination. Baxter et al., showed in his study that persons >65 years old with a > 7.5% chance of death in the upcoming year are less likely to receive the influenza vaccine. Because persons who are most likely to die are less likely to receive the vaccine, vaccination appears to be associated with a much lower chance of dying; thus, the “effectiveness” of the vaccine is in great part due to the selection of healthier individuals for vaccination, rather than due to true effectiveness of the vaccine. Physicians would elect to vaccinate the healthier co-morbid patient rather than the frail co-morbid patient. The curvilinear relationship suggested by Baxter “implies” that increasing illness means increasing vaccination up to a point, and then, as people come closer to the end of life, there is a decrease in vaccination coverage, which seems reasonable in our study.²⁴

Furthermore, healthier people may decline the vaccination if offered due to their belief in a lack of benefit in view of their well-being. In a study²⁵ to assess the knowledge and beliefs of family physicians in Lebanon regarding influenza and pneumococcal vaccines, Lebanon adult vaccination was not optimal; and barriers for recommending vaccines were reported by 14 participants (32.5%). Concerning the influenza vaccine, the most commonly reported barriers to vaccinations were the patient’s refusal and non-availability of the vaccine, followed by physicians’ concerns about efficacy and safety.²⁵ Thus, when faced with a reluctant patient or a patient who refuses vaccination, physicians who have concerns about safety would find it easier to justify their non-adherence to recommendations and blame it on the patients’ hesitancy and refusal.

If we consider the discrepancy in cardiac events between vaccinated and unvaccinated patients, a marker of effectiveness of the vaccine,²⁶ our study confirms that selection bias greatly confounds the analysis of observational studies of influenza vaccine effectiveness in elderly patients as previously reported.^27,28 A crucial but inevitable flaw in observational studies is the presence of systematic differences between vaccine recipients and non-recipients.^27,28

When looking at both groups in view of their risk factors, vaccinated patients with risk factors for the cardiac disease were not significantly different than non-vaccinated with no risk factors in terms of having cardiac events (OR = 0.97 vs. 1.74). Interestingly, the occurrence of HD in the presence of vaccination revealed a non-significant decrease trend with the increase in a number of risk factors (OR = 1.61 vs 0.97).

In a large population-based study, influenza vaccination in people with type 2 diabetes was associated with reductions in rates of hospital admission for acute cardiovascular and respiratory diseases, and in all-cause mortality across seven influenza seasons. Influenza vaccination was associated with lower rates of admission to hospital for acute MI, but this association was not statistically significant as in our study.²⁹ The protective effect of the influenza vaccination seems to be due to the stabilization of the plagues as was shown in vaccinated mice with influenza. So probably the co-morbid patients in this study were protected due to the same mechanism of plaque stabilization. Following the same token, Bermúdez-Fajardo et al. showed that in mice immunized for Influenza, vaccination was shown to promote stable atherosclerotic plaques and generates athero-protective immune responses in apoE−/− mice. Vaccination of mice with a high dose of one brand of Influenza vaccine (which is the same dose recommended by the manufacturer to vaccinate humans adults for influenza) developed a smaller atherosclerotic lesions with lower lipid content but richer in smooth muscle cells and collagen. The authors also showed that the vaccine was capable to lower plasma levels of interferon gamma (IFN), interleukin (IL)-2 and tumor necrosis alpha (TNF). Despite the need to prove a similar immune response in humans, these findings pose a reasonable biologic plausibility.³⁰

Strength

The strength of this study is that patients were unselected and represent the real-life situation.

The patients were followed up only in this clinic following the managed care model and thus all treatments and patient information were well documented and centralized in one center. Despite being a retrospective study, we did not encounter missed data that would have affected the size of the power.

Limitations

Besides some of the biases mentioned above and which could have affected the results, there are other limitations that are worth consideration. The study was conducted in one private center and included a relatively a small sample size that failed to detect any significance between the experimental and control group. This calls for future studies that need to look at larger cohorts in more than one center across the country. Another important limitation was the lack of data on family history of heart disease, a major contributor to cardiac events that could have skewed the results. It is reasonable to assume that physicians vaccinated the more sick and co-morbid but it would have been more authentic had the decision been based on a national policy which is currently lacking in Lebanon. Leaving the decisions to the physician to decide on influenza vaccination would be challenged by many barriers and having a policy in place will help standardize and organize the influenza vaccination for the high-risk target groups.

Even though developed countries have recommendations in place for annual influenza vaccination, the vaccine coverage rates vary widely between the elderly and others for whom the vaccine is recommended. In a recent survey conducted in Europe over three influenza seasons (2013–2014, 2014–2015, and 2015–2016) the vaccine coverage rates did not reach the 75% target set by the WHO. In fact, the survey showed decreasing vaccination rates in the elderlies in most of the European countries posing a major health challenge even in countries where policies are available.³¹ Thus, besides the installment of polices, continuous surveys to monitor vaccination coverage and adherence to policies need to be revisited. The data of our study sets in to further support such policies.

Conclusion

Our results suggest that vaccination for influenza may contribute to protection against HD in patients with multiple risk factors for heart diseases. The protective effects are more emphasized with an increase in the number of comorbid conditions, however not significant. The potential effects of the vaccine may be related to the protection against influenza.

Disclosure of potential conflicts of interest

The authors declared no conflicts of interest.

Additional information

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.