Understanding the Flu: Host and Environmental Factors Associated With Susceptibility to Influenza Virus-Induced Disease
The relationship between smoking and increased risk for many cancers and chronic diseases is well known. The risk of respiratory infections is also increased among both active and passive smokers (124). The effect of smoking on the risk of respiratory infection is mediated through both structural changes in the respiratory tract and alterations in cellular and humoral immune function (124). Although many studies have investigated the effect of smoking on respiratory infections, there are only a handful of reports that have specifically examined influenza virus infection in humans. Finklea et al. (125) conducted a study among 1811 male cadets at the Citadel during the influenza A epidemic in 1968. An increased incidence of self-reported influenza correlated with increased smoking activity. Heavy smokers and light smokers had 21% and 10% more influenza than non-smokers, respectively. Increased smoking was not found to be associated with increased severity of disease.
A sero-survey of the cadets showed that smokers had evidence of increased subclinical influenza infection (those who did not report symptoms but were seropositive). Two studies done in Israeli military bases have also examined smoking status and influenza susceptibility. The first was done among 176 female military recruits during an outbreak of influenza A virus in 1979 (126). Smoking status and the incidence of influenza-like illness (ILI) was assessed by questionnaire. There was a higher incidence of ILI in smokers (60%) compared to nonsmokers (41.6%). Illness was significantly more severe in smokers and the risk of ILI attributable to smoking was 13%. In the second study, the incidence of influenza among 336 men in an Israeli military unit was assessed by self-report and by serologic survey (127). There was an increased incidence of clinically apparent influenza among smokers (68.5%) compared to nonsmokers (47.2%). In addition, the severity of influenza, measured by the percentage that had lost work days or required bed rest, was higher among smokers (50.6%) than nonsmokers (30.1%). The attributed risk due to smoking was 31.2% for all cases of influenza and 40.6% for severe influenza. The results from the serologic survey showed that there was only a small increase in anti-influenza antibody levels in smokers compared to nonsmokers.
A study examining the effects of smoking on responses to influenza virus vaccination in a cohort in Western Australia found that there were no differences in hemagglutination-inhibiting (HI) antibody titers and longevity between healthy smokers and non-smokers following live-attenuated influenza virus infection (128). However, HI longevity was significantly reduced in smokers following vaccination with a killed subunit influenza virus vaccine when compared with non-smokers (P<0.05). In addition, the same study reported that in individuals that were exposed to an epidemic strain of influenza virus, smokers with little to no pre-epidemic HI antibody were significantly more susceptible to epidemic influenza (36.5%) than non-smokers (21%). Noah et al (129) conducted a cohort study among healthy young smokers, non-smokers exposed to secondhand smoke, and non-smokers to examine the effects of smoking on nasal mucosal responses after being infected with a live-attenuated influenza virus vaccine strain composed of influenza A H3N2, influenza A H1N1 and influenza B strains. Nasal lavages from smokers had increased viral RNA and suppressed IL-6 response compared to non-smokers, with those exposed to secondhand smoke having an intermediate response. Given that these studies were all performed in healthy, young, physically fit populations, the results may not be generalized to the population-at-large, especially the aged and those with chronic underlying health conditions. However, these studies do suggest that smoking contributes to increased susceptibility to influenza and potentially decreased immunological responses following influenza vaccination.
Studies using in vitro and mouse models of influenza virus infection have also investigated the effects of cigarette smoke on the innate and adaptive immune responses during infection. Wu et al (130) found that following exposure of human lung organ cultures to cigarette smoke extract (CSE) and then infection with H1N1 influenza A (PR8) virus, influenza virus-induced IP-10, IL-8, and MCP-1 expression was significantly decreased when compared with infected cultures that were not exposed to CSE. Moreover, the CSE-mediated suppression of innate antiviral responses was found to be primarily RIG-I dependent, although the expression of other pattern recognition receptors such as TLR3 was also suppressed by CSE exposure during influenza virus infection. Another study investigating the effects of cigarette smoke on pulmonary T cell responses during influenza A virus infection found that T cell mediated IFN-γ production was decreased in C57BL/6 mice that were exposed to cigarette smoke during influenza A (PR8) virus infection (131). Moreover, mice exposed to cigarette smoke showed enhanced viral burden and significantly increased morbidity (P = 0.006) and mortality (P = 0.008) following influenza virus infection when compared to control mice. Horvath et al (132) found that differentiated nasal epithelial cells (NECs) from smokers exhibited significantly decreased influenza A virus-induced, interferon-related proteins following infection when compared with NECs from non-smokers. Following co-culture of monocyte-derived dendritic cells (mono-DCs) with NECs from smokers and non-smokers, mono-DCs co-cultured with smoker NECs also showed reduced interferon-induced protein expression compared with mono-DCs co-cultured with non-smoker NECs during influenza virus infection. These studies support the human cohort studies in suggesting that exposure to cigarette smoke can impair immunological responses to influenza virus infection and lead to increased susceptibility to influenza virus infection.
Although chronic alcohol consumption is known to be associated with increasing susceptibility to and severity of respiratory bacterial infections, there has been very little research on the effect of alcohol on viral respiratory infections, including influenza (133-135). Two experimental laboratory studies have been done on this topic, one on chronic alcohol consumption and the other on fetal alcohol exposure. Meyerholz et al (136) conducted a study in C57BL/6 and BALB/c mice in which chronic ethanol (EtOH) exposure was simulated by gradually increasing the concentration of the drinking water to 18-20% over a 4 or 8 week period, after which they were infected with influenza intranasally with H1N1 or H2N2 influenza virus strains. Increased mortality (50% vs. 0%) and morbidity was observed among EtOH-exposed mice. Lungs from EtOH-exposed mice had more inflammation, edema, and consolidation, higher virus titers and a prolonged presence of viral antigen in the lung. There were decreased numbers of total and influenza-specific CD8+ T cells in EtOH-exposed mice, which was worsened in mice with longer periods of EtOH exposure.
A follow-up study showed that treatment with oseltamivir reduced morbidity and mortality in mice exposed to chronic EtOH (137). A study by McGill and colleagues (138) evaluated the effect of fetal and chronic alcohol exposure on the severity of influenza A virus infections. Fetal alcohol exposure (FAE) was simulated by exposing the pregnant C57BL/6 or BALB/c female to 10-12% EtOH during gestation and until pups were weaned. Of the FAE-exposed pups, 50% were exposed to 20% EtOH beginning at 5-7 weeks of age for a duration of six weeks to simulate adult chronic alcohol exposure (FAE + chronic). Mice in each of the exposure groups were then infected with H1N1 or H2N2 influenza virus strains when they reached adulthood. An increase in mortality, increased lung viral titers and reduced viral clearance was observed in FAE mice, and these effects were exacerbated in FAE + chronic mice. Total and influenza virus-specific CD8+ T-cells were reduced in FAE mice and further reduced in FAE + chronic mice. In addition, there were reduced numbers of B-cell foci in the lungs and reduced influenza virus-specific serum antibodies in FAE and FAE + chronic mice. The humoral response in FAE + chronic mice was severely diminished, suggesting that alcohol exposure diminishes the adaptive immune response to influenza virus infection.
Air pollution is known to have many adverse effects on health, particularly on respiratory health. In addition to increasing morbidity and mortality associated with chronic respiratory illnesses such as asthma and COPD, air pollution also affects susceptibility to respiratory viral infections (139). Several epidemiologic and experimental studies have been conducted to evaluate the effect of different types of air pollutants on influenza susceptibility. One of the earliest studies was conducted during the 1968 Hong Kong influenza epidemic, among schoolchildren in California from five different communities (140). Each of the communities was assigned as low, intermediate or highly polluted based on daily maximum hourly oxidant averages collected by air monitors. The incidence and duration of ILI and other febrile illnesses was assessed by parental questionnaire, and a serosurvey was conducted at the end of the influenza epidemic. The authors found that 21.5% of the students had febrile clinical influenza as reported by the parents and 24.6% had elevated convalescent serum antibodies to influenza virus. The incidence of influenza (both clinical illness and those with anti-influenza virus antibodies) was lowest in the two most polluted cities. An ecologic study on acute effects of air pollution on the incidence of influenza was investigated during a 1974 to 1975 epidemic of influenza in Sofia, Bulgaria (141). The authors found that the atmospheric concentrations of nitric oxides, formaldehyde, and sulfur dioxide were positively correlated with the number of influenza cases, while the concentration of oxidants had an inverse correlation. The effect of nitric oxides and formaldehyde appeared to be lagged by several days, while sulfur dioxide induced an effect on the next day, and the effect of oxidant exposure was seen on the same day.
Another ecologic study used a time series analysis to evaluate the effect of sulfur dioxide pollution on influenza incidences between 1976 and 1987 (142). Adjusting for seasonal and temperature effects, the authors found a significant association between hourly sulfur dioxide levels and weekly influenza incidence, with a lag between exposure and incidence of approximately two weeks. Wong et al. (143) also conducted an ecologic time series study to examine the effect of four different air pollutants on hospital admissions for respiratory and cardiovascular diseases in Hong Kong. Analysis of the admissions for pneumonia- and influenza-related illnesses revealed that there were significantly increased admission rates for every 10 μg/m3 increase of nitrogen dioxide, ozone, and PM10 (particulate matter of 10 microns or less). In a double-blind, placebo-controlled, randomized trial conducted by Goings et al. (144), subjects inhaled nitrogen dioxide for 2 hours per day for three consecutive days, followed by intranasal challenge with a live-attenuated strain of H3N2 influenza A virus. Only one infected person developed symptoms, but subjects exposed to 1-2 ppm of nitrogen dioxide had increased infection rates compared to controls (91% vs. 71%), although this difference was not statistically significant.
9b. Diesel exhaust
Diesel exhaust (DE) is a mixture of particulate matter and gases made up of hundreds of organic and inorganic compounds (145). Diesel exhaust may represent a large proportion of particulate matter air pollution in urban areas, and exposure to DE is known to induce pulmonary inflammation (145). To date, only one study has examined the effect of inhaled diesel exhaust on influenza infection in human volunteers. Noah et al. (146) conducted a double-blind randomized-controlled study among subjects with or without allergic rhinitis and exposed them to either clean air or diesel exhaust for 2 hours, followed by intranasal inoculation with the live-attenuated influenza A virus vaccine commonly known as FluMist. Diesel exposure resulted in a significant increase in IFN-y response in both allergic and non-allergic subjects. Viral RNA levels were significantly higher in those exposed to diesel exhaust, and this effect was worse in allergic subjects. Several experimental laboratory studies have also focused on the effects of exposure to diesel exhaust on susceptibility to influenza. Jasper et al. (147) examined the effect of aqueous DE particles on H3N2 influenza A virus infection in human A549 respiratory epithelial cells. Exposure to aqueous DE increased the total numbers of infected cells because of increased virus attachment and uptake. In addition, DE-exposed cells exhibited evidence of oxidative stress as well as enhanced viral replication and type I IFN production. A study in mice exposed to DE showed that BALB/c mice exposed to low doses of DE (0.5 mg/m3 DE for 4 hours per day for five consecutive days) were more susceptible to H3N2 influenza A virus than mice exposed to a higher dose (2.0 mg/m3) for the same duration (148). The authors attributed the increased susceptibility to the decreased expression of antimicrobial surfactant proteins that was observed in mice exposed to the low dose of DE but not in those exposed to the higher dose.
In contrast to the acute effects of DE exposure in the study by Ciencewicki et al, chronic exposure to DE and coal dust was evaluated in a study by Hahon et al (149). CD-1 inbred mice were subjected to chronic exposure to coal dust (CD), diesel exhaust (DE), or both (CD+DEE) at 2.0 mg/m3 for one, three or six months, followed by infection with influenza A (PR8) virus by aerosol. No differences in mortality were observed in any of the groups of mice for all durations of exposure. Increased disease severity was observed in mice exposed to DE and CD+DE for 3 or 6 months, as evidenced by increased viral titers and lung consolidation, and decreased serum antibodies. Mice exposed to both DE and CD did not show any appreciable differences in severity of disease beyond what was observed in DE only-exposed mice, indicating that the increased severity was mediated by diesel exhaust and not by coal dust. A study examining the effects of silica dust, which is often a component of coal dust, had similar results wherein there were no differences in influenza-associated mortality or viral antibody titers in mice with short or long term exposure to silica dust (150).
Ozone is one of the most common oxidant pollutants in ambient air and is the principal component of smog (151). Ozone exposure is known to have many adverse health effects on the respiratory system as well as the immune system (152). In an experiment by Wolcott et al (153), the effect of exposure to 0.5 ppm ozone prior to WSN strain influenza A virus infection and after infection was evaluated in Swiss-Webster inbred mice. Exposure to ozone after infection reduced mortality and increased survival time, and this effect was independent of whether mice had been exposed to ozone prior to infection. The reduction in mortality appeared to be due to less widespread infection in the lungs of ozone-exposed mice because ozone exposure had no effect on viral titers or on serum and lung neutralizing antibody titers. Selgrade et al (154) exposed CD-1 inbred mice to 1 ppm of ozone for three hours per day for five days, and infected groups of mice with H3N2 influenza A virus on each of the five days of exposure. Only mice infected on day two of ozone exposure showed a significant increase in mortality and decrease in mean survival time compared to control mice, although no differences in mortality were seen when ozone exposure was reduced to 0.5 ppm. These mice also had increased morbidity as evidenced by increased wet lung weights, enhanced lung pathology, and decreased pulmonary function. The authors suggested that the effects of ozone enhanced the inflammation and edema associated with influenza virus infection. A study by Jakab and Hmieleski (155) examined the effect of 0.5 ppm ozone exposure following aerosol infection with influenza A (PR8) virus in mice. Ozone-exposed Swiss inbred mice had reduced severity of influenza virus infection as evidenced by less widespread infection in the lung, although ozone exposure had no effect on viral titers. Lower serum antibodies and reduced T and B lymphocytes in the lungs of ozone-exposed mice indicated that ozone exposure impaired the immune response to infection. A follow-up study in which mice were subjected to chronic ozone exposure after infection found that ozone exposure reduced the severity of acute lung injury caused by influenza virus infection, but long term exposure enhanced residual damage (156). The authors proposed that this occurred through ozone-mediated inhibition of the repair process.
9d. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is a by-product of combustion of fossil fuels, wood, and municipal and industrial waste (157). Laboratory studies in inbred mice and rats have shown that exposure to TCDD is immunotoxic, resulting in influenza-related mortality in animals exposed to sublethal doses of H3N2 influenza A virus (158,159). The mechanism for this effect is still unclear and is the subject of ongoing research.Continued on Next Page »