Advertisement
Research Article

The Impact of Weather on Influenza and Pneumonia Mortality in New York City, 1975–2002: A Retrospective Study

  • Robert E. Davis mail,

    red3u@virginia.edu

    Affiliation: Department of Environmental Sciences, University of Virginia, Charlottesville, Virginia, United States of America

    X
  • Colleen E. Rossier,

    Affiliation: Department of Environmental Sciences, University of Virginia, Charlottesville, Virginia, United States of America

    X
  • Kyle B. Enfield

    Affiliation: Division of Pulmonary and Critical Care, Department of Medicine, University of Virginia Health System, Charlottesville, Virginia, United States of America

    X
  • Published: March 28, 2012
  • DOI: 10.1371/journal.pone.0034091

Abstract

The substantial winter influenza peak in temperate climates has lead to the hypothesis that cold and/or dry air is a causal factor in influenza variability. We examined the relationship between cold and/or dry air and daily influenza and pneumonia mortality in the cold season in the New York metropolitan area from 1975–2002. We conducted a retrospective study relating daily pneumonia and influenza mortality for New York City and surroundings from 1975–2002 to daily air temperature, dew point temperature (a measure of atmospheric humidity), and daily air mass type. We identified high mortality days and periods and employed temporal smoothers and lags to account for the latency period and the time between infection and death. Unpaired t-tests were used to compare high mortality events to non-events and nonparametric bootstrapped regression analysis was used to examine the characteristics of longer mortality episodes. We found a statistically significant (p = 0.003) association between periods of low dew point temperature and above normal pneumonia and influenza mortality 17 days later. The duration (r = −0.61) and severity (r = −0.56) of high mortality episodes was inversely correlated with morning dew point temperature prior to and during the episodes. Weeks in which moist polar air masses were common (air masses characterized by low dew point temperatures) were likewise followed by above normal mortality 17 days later (p = 0.019). This research supports the contention that cold, dry air may be related to influenza mortality and suggests that warning systems could provide enough lead time to be effective in mitigating the effects.

Introduction

It is well known that intra-annual mortality exhibits a pronounced winter peak in locations with seasonal climates [1], [2]. In the United States, mortality arising from respiratory disease is 50% higher in winter than in summer [1]. One contributor to this excess winter mortality is influenza [3], the timing and severity of which can vary significantly from year to year. The association between the annual influenza peak and winter weather in temperate locations has lead to the hypothesis that weather variability could influence influenza mortality variations [4], [5].

Recent research examining climatic influences on influenza transmission has galvanized interest in this topic. Airborne transmission of Influenza A/Panama virus between guinea pigs was more likely at low temperatures and relative humidities [4]. Because temperature is physically/mathematically linked to relative humidity and thus confounds interpretation, a mass-based humidity measure (such as specific humidity or vapor pressure) provides a stronger relationship to influenza transmission [5]. Given the cold-season decline in specific humidity in temperate climates, a retrospective study [6] showed an association between year-to-year humidity variations and the timing of the influenza seasonal onset in U.S. states. This differs from tropical locations which generally lack a seasonal influenza peak [7]. One theory is that tropical climates are dominated by direct contact transmission which, unlike airborne transmission, is not influenced by air temperature and humidity [7].

These recent studies [4], [5], [6], [7] have motivated several reviews on climate and influenza seasonality [8]. [9], [10]. These reviews, which approach the issue from atmospheric sciences, virological, and epidemiological perspectives, do not reach firm conclusions on the causes of influenza seasonality but suggest that those causes are complex and multifactorial and that the solution will require interdisciplinary cooperation.

A variety of theories exist as to how weather and climate might exert some influence on influenza seasonality. Low temperatures enhance viral stability [4], reduce mucosal blood flow [11], and/or diminish mucociliary clearance [12]. Correlations exist between the number of upper respiratory infections and cold-air outbreaks [13]. It is commonly assumed that winter indoor crowding enhances influenza virus transmission, though direct evidence is lacking [8]. Low humidity conditions, which are often but not always accompanied by low temperatures, enhance survival times of viral aerosols [14], [15], [16]. If correct, these relationships suggest that indoor winter heating without humidification could enhance influenza transmission [8], as indoor absolute humidity tends to be correlated with outdoor values and thus is typically lower in winter [6].

From a micro-physical perspective, there is evidence that both droplet size and transmission mode depend on ambient environmental factors [17], [18]. Small droplets can remain airborne longer whereas large drops tend to precipitate, suggesting that ambient conditions would influence whether airborne- or contact-mode transmission predominates. The virulence of the influenza virus changes constantly as the virus undergoes antigenic shift and antigenic drift [8]. Although virulence depends upon a variety of factors, one possible atmospheric influence is that cold, dry air allows the lipid envelope encasing the influenza virus to remain intact longer, increasing the likelihood of infection [19].

Theories proposing that factors other than weather/climate are responsible for influenza seasonality include cycles in viral interference [20] and intrinsic temporal viral dynamics that operate independent of external forcing factors [21]. Some research suggests that the seasonality is driven by the school calendar in which students reconvene after a summer break, but this timing is not consistent with the typical onset of influenza in early winter [22].

We examine the hypothesis that cold and/or dry weather enhances human pneumonia and influenza (P&I) mortality through a retrospective study of daily mortality in New York City and environs from 1975–2002. We hypothesize that periods with colder and/or less humid conditions exhibit excess P&I mortality for a period of time following those climatic conditions. Our research differs from recent work on this topic that examined the timing of the influenza season onset over large geographic areas [6]—our focus is on the influence of daily weather on influenza characteristics for a single, large metropolitan area.

We selected New York City for this study for several reasons. Our study examines daily mortality, and statistical robustness is enhanced when the daily sample size is sufficiently large. Weather obviously has a high spatial variability, so it is important that the observed weather be representative of the environmental conditions likely experienced by the decedents. In addition, New York City's mid-latitude location provides a high degree of both interannual and intra-annual variability in weather and climate, so this variability provides a wider range of sample conditions. Thus, New York City has both a large enough population to provide a consistent daily mortality signal while the population density is high enough that the weather observed at a single station is sufficiently representative of conditions experienced throughout the metropolitan area.

Methods

Outcome Data

We conducted a retrospective cohort study of pneumonia and influenza (P&I) mortality of residents of the New York City metropolitan area. Daily frequencies of P&I mortality were tallied from National Center for Health Statistics archives for the New York City Consolidated Metropolitan Statistical Area (which, as defined in the year 2000, includes 30 counties in New York, New Jersey, Connecticut, and Pennsylvania). This period of record spans three revisions of the International Classification of Diseases (ICD) codes (Table 1).

thumbnail

Table 1. International Classification of Diseases periods and codes for pneumonia and influenza.

doi:10.1371/journal.pone.0034091.t001

In the National Center for Health Statistics mortality files that we used for this research, all information that could allow an individual to be identified has been removed. This research utilized only mortality counts for a large metropolitan area. These de-identified counts are stored in governmental archives for the purposes of retrospective research; because all personal identifying information is redacted, consent is not required. Thus, this research is exempt from IRB review under the auspices of Title 45 Part 46 exemption category 4.

Pneumonia or influenza must be listed as the primary cause of death to be included in this analysis. These diseases are commonly combined as an endpoint because of specific challenges associated with influenza. First, the number of deaths attributable to influenza is difficult to estimate directly because of a lack of virologically-confirmed infections. Second, many influenza-associated deaths occur from secondary complications when influenza viruses are no longer detectable by laboratory means [6], [23], [24], [25]. Third, the use of P&I mortality in the study of influenza also reflects the problems associated with other measures of prevalence of influenza. Reporting of cases of influenza through routine channels is unsatisfactory because mild influenza may be under-diagnosed and the use of laboratory confirmation is skewed by the impact of variable testing patterns based on prevalence of disease [26]. Finally, multiple studies have shown that there exists a relationship between influenza morbidity and P&I mortality that can be mathematically described and used in epidemiological studies [3], [27], [28], [29], [30], [31].

Daily deaths were aggregated into ten age groups (0–4, 5–14, 15–24, 25–34, 35–44, 45–54, 55–64, 65–74, 75–84 and >84) and standardized via direct standardization [32] based upon the age distribution of population of the United States in the year 2000. This procedure adjusts for temporal changes in age demographics over the period of record based on data for each county using U.S. Census archives, thereby allowing for consistent comparisons of mortality rates over time. After age standardization, the long-term mean for each day (within ICD period) was subtracted from that day's standardized P&I mortality to remove the inherent seasonal signal from the time series. Because our goal was to determine if cold and/or dry days or periods within a year were related to P&I mortality peaks, this level of granularity requires the examination of daily data with the seasonal signal removed.

Examination of the P&I mortality time series exhibits obvious temporal discontinuities that are exactly coincident with the dates of ICD revision code changes. We removed this artifact by converting each day's P&I mortality to a z-score by dividing the mean departure by the standard deviation separately for each of the three relevant ICD periods (Table 1). The time series of age-standardized mortality prior to and after deseasoning and z-score adjustment is shown in Figure 1a and Figure 1b, respectively. Because of the general lack of influenza in the summer months, June, July, and August were removed from the analysis as they had the potential to distort any relationships during the primary influenza season.

thumbnail

Figure 1. Time series of pneumonia and influenza mortality for New York City from September, 1975–May, 2002.

a) (top) Daily age-standardized pneumonia and influenza mortality time series (deaths per million; June, July and August have been deleted). The relevant periods for the International Classification of Diseases (ICD) are identified by a thick vertical line; b) (bottom) Resulting mortality time series after removing the seasonality and converting to z-scores for each ICD period. Vertical dividers identify influenza seasons (September–May) with the year assigned to the January–May period (i.e., December, 1979 is in the 1980 flu “season,” labeled as “80” on the x-axis).

doi:10.1371/journal.pone.0034091.g001

Mortality data were smoothed using a 17-day leading moving average (e.g., mortality on January 1 is the mean from January 1–17). This smoother was selected after testing a variety of filter lengths and based upon prior research [6].

We examined both daily P&I mortality “events” and longer mortality “episodes.” “Events” are days with (smoothed) mortality at least one standard deviation above the long-term (smoothed) mean for that date. This z≥1 criterion was chosen because the frequency distribution of smoothed mortality is positively skewed with the tail beginning at approximately one standard deviation. After smoothing, there is an obvious tendency for high mortality events to cluster into prolonged periods when the z≥1 threshold is exceeded (Figure 2). We thus identified 12 P&I mortality episodes over the 28-year period of record (Table 2).

thumbnail

Figure 2. Identification of 12 mortality “episodes” that exceeded the z≥1 criterion.

The time series is Figure 1b smoothed with a 17-day centered moving average filter. A centered smoother is used here to more clearly present the peak times of the mortality episodes (see Table 2).

doi:10.1371/journal.pone.0034091.g002
thumbnail

Table 2. Characteristics of the 12 high mortality “episodes.”

doi:10.1371/journal.pone.0034091.t002

To summarize the outcome data treatment, after age standardization the daily mortality data were deseasoned to remove the large influence of season on respiratory infection and converted to z-scores to adjust for discontinuities related to ICD coding. These data were then smoothed using a 17-day leading smoother to account for the inherent lag between infection and mortality. Days or periods with z-scores>1 were identified as mortality “events” or “episodes,” respectively (Figure S1).

Weather Data

Hourly climate data from La Guardia Airport, New York, were retrieved from National Climatic Data Center archives. We utilize 1200 and 1900 Universal Time Coordinate (UTC) air temperature (T) and dew point temperature (Td) to approximate the typical times of the warmest and coldest hours of the day (7 or 8 a.m. and 2 or 3 p.m. local time). The dew point temperature is the temperature at which water vapor begins to condense via cooling at constant pressure. We use dew point as a measure of the amount of moisture in the air because, unlike relative humidity, it is independent of air temperature [5].

In addition to dew point temperature, we employ an air mass classification, which has the advantage of incorporating a variety of weather variables into a single nominal variable. Specifically, we utilize the Spatial Synoptic Climatology (SSC) [33] which classifies each day's weather into one of six air mass types [dry moderate (DM), dry polar (DP), dry tropical (DT), moist moderate (MM), moist polar (MP), moist tropical (MT)]. A seventh transition (TR) category identifies days characterized by a significant change in weather—typically a frontal passage [34]. The SSC uses air temperature, dew point temperature, wind, surface air pressure, and total cloud cover observations taken four times per day as input to the classification. Thus, the SSC is a multivariate nominal classification of daily weather conditions. The approach has been utilized in a variety of human health applications [35], [36], [37], [38], [39], [40], [41].

The temperature and dew point time series were converted to z-scores to remove seasonality and then smoothed using a centered 5-day moving average filter. This filter length was employed after examining various options because it represents a balance between high frequency weather events and more long-term (monthly to seasonal) trends.

The seasonality of the SSC air mass types was removed by comparing the presence (coded 1) or absence (coded 0) of each air mass type on each day of the year to the long-term average frequency. For example, if Moist Moderate air was present on average 30% of the time on January 1 over the period of record, then its occurrence on January 1, 2000 would result in a value of +0.7 for that day. These daily anomalies were then converted into continuous variables using a centered 7-day moving average filter for each SSC category.

In summary, raw dew point observations were first de-seasoned by conversion to z-scores and then smoothed using a 5-day filter. Daily air mass frequencies were converted from a nominal to a continuous variable by first adjusting for the long-term frequency on each day and then smoothing those frequencies using a 7-day moving average (Figure S2).

Statistics

Daily Analyses.

A series of t-tests were employed to determine if temperature, dew point, and air mass frequency differed between high P&I mortality events (z≥1) and non-events. To address the temporal autocorrelation in the weather variable time series and the resulting overestimate of the true degrees of freedom, the sample size was adjusted based upon the lag one temporal autocorrelation [42] (Wilks 2006) as follows:
where N = number of observations

 N′ = adjusted degrees of freedom

 P = lag one temporal autocorrelation.

N′ was then adjusted again based on the length of the smoother employed to determine the final effective sample size. For these and all other tests, a Type I error rate of 0.05 was employed and Levene's test for equality of variances was used to determine if pooling of the samples was required.

The following tests were performed:

  1. smoothed temperature, dew point, and air mass frequency, lagged 17-days, between mortality events (z≥1) and non-events (z<1) using an unpaired two-sample t-test (Figure S3);
  2. same as in 1 for unsmoothed temperature and dew point temperature (to determine if a strict 17-day lag exists); and
  3. same as in 1 but using a one-sample t-test (to account for the possible influence of disparate sample sizes between groups).
Episodic Analyses.

For each high P&I mortality episode, we calculated the duration (in days), the summed total mortality over the entire episode, and the average daily episode mortality (Table 2). These quantities served as dependent variables in a linear regression analysis vs. the independent (weather) variables (1200 and 1900 UTC T and Td and the deseasoned SSC frequencies). We used a 17-day lag in which the mean for each variable was calculated across the days in the episode and the preceding 17 days. This lag was selected based upon prior research that examined absolute humidity [6] and the results of other studies [28], [43].

Given the relatively small number of episodes, we used bootstrapped regression analysis to generate a robust estimate of the regression coefficients. Based on the initial full sample, data sets of the same size were generated by randomly sampling variable pairs, with replacement, and estimating the regression parameters from that sample using ordinary least-squares. This procedure was repeated 10,000 times and the resulting suite of regression coefficients was examined to determine if the 2.5 percentile and 97.5 percentile observations were of the same sign. If so, the regression slope was deemed to be statistically significant [44].

Results

Daily Analyses

In the daily analysis, 1200 UTC dew point temperature was significantly lower for events than for non-events (p = 0.003), a result that is consistent with the hypothesis that drier conditions are related to enhanced P&I mortality. However, 1900 UTC dew point was higher during events (p = 0.036), a result that contradicts the underlying hypothesis.

When this test was repeated without smoothing the weather variables, only 1200 UTC dew point was significant (p = 0.028; Table 3). The lack of a relationship for 1900 UTC dew point (p = 0.181) suggests that the finding of high afternoon dew points during mortality events using smoothed weather data was not robust.

thumbnail

Table 3. Results of t-tests comparing weather variables between mortality event days to non-event days.

doi:10.1371/journal.pone.0034091.t003

Because the high number of non-event days vs. event days can bias the t-test, an additional test was performed comparing the 5-day smoothed weather variables during events to the long-term mean. Here, 1200 UTC temperature and dew point were both significantly lower during high P&I mortality events (Table 3).

For the daily SSC analysis, lower frequencies of moist moderate (MM) air (p = 0.019) and higher frequencies of moist polar (MP) air (p<0.001) occurred 15–19 days before high mortality events (Table 3). No relationship was found for dry polar (DP) air (p = 0.277).

Episodic Analyses

Temperature, dew point temperature, and air mass frequencies were examined 17 days prior to and throughout each of the 12 high P&I mortality episodes identified from 1975–2002. There is a statistically significant negative relationship between episode duration and mean 1200 UTC dew point (r = −0.61, p<0.05; Figure 3a). The longest episode (57 days in 1981) was associated with a dry period during which the mean dew point was more than 0.8°C below normal for that time of year. Total episode mortality is likewise negatively correlated with 1200 UTC dew point (r = −0.56, p<0.05; Figure 3b). Two of the three episodes with the lowest mortality also had dew points that were near normal, whereas the higher mortality episodes exhibited drier conditions. No significant relationships were found for afternoon variables, air temperature, or any of the SSC air mass types.

thumbnail

Figure 3. Scatter plots of pneumonia and influenza mortality episode duration and total mortality vs. dew point temperature.

a) (left) Total episode duration (days) vs. mean episode dew point temperature (°C) (r = −0.61). b) (right) Total episode mortality (in z-score units) vs. mean episode dew point temperature (°C) (r = −0.56). The regression line shown in both graphs is for the least squares linear regression of the full data set. Both of these relationships were determined to have statistically significant slopes based upon 10,000 bootstrapped samples.

doi:10.1371/journal.pone.0034091.g003

Discussion

Dew point temperature is commonly used by atmospheric scientists to measure humidity because it is relatively invariant to pressure and temperature changes and thus is a conservative quantity. In New York City from 1975–2002, periods with high P&I mortality were preceded 2–3 weeks by periods with low morning dew points. Furthermore, for the 12 high mortality episodes identified in that period, morning dew point was negatively correlated with both episode duration (r = −0.61) and total episode mortality (r = −0.56). This finding of a linkage between dry air and influenza mortality is consistent with the results of recent research [6] showing that absolute humidity (a correlate of Td) influences the timing of the onset of the influenza season in various U.S. states (including New York state). Our results provide some limited evidence supporting laboratory studies linking dry air to higher airborne infection rates among guinea pigs [4].

The association between high frequencies of Moist Polar air masses prior to high mortality events is consistent with the dew point results. The average morning dew point in Moist Polar air in New York City is lower than for any air mass other than Dry Polar (Table 4). Although it may seem counter-intuitive that a “moist” air mass has a low dew point temperature, dew point must (physically) be less than or equal to air temperature, so cold air masses typically have low dew points, especially in winter. In general, Moist Polar air masses are relatively uncommon in the cold season in New York, occurring only 7.3% of the time (Table 4) and are associated with cold, overcast and often stormy conditions with moist air arriving from the Atlantic Ocean [33].

thumbnail

Table 4. SSC air mass frequencies, mean 4 a.m. dew point temperatures, and ranks for October–March for New York City (La Guardia).

doi:10.1371/journal.pone.0034091.t004

We also identified a significant relationship between low Moist Moderate air mass frequencies 3–4 weeks before mortality episodes. In an effort to understand this association, we calculated the correlation between Moist Moderate frequencies and the other air mass types during that period. Moist Moderate is negatively correlated with the two driest air masses—Dry Polar (r = −0.47) and Moist Polar (r = −0.31) (Table 4). Thus, the significant association between low Moist Moderate frequencies prior to P&I mortality events appears to be a proxy for the cool, low dew point conditions that are common when Moist Polar and Dry Polar air masses are present. On average, cold season dew points are 5.5°C higher in Moist Moderate air masses than in Moist Polar (Table 4).

The lack of a direct Dry Polar relationship is surprising, as Dry Polar air masses exhibit the most extreme combination of cold air and low humidity. Dry Polar is far more common than Moist Polar in New York winters, however, so its high daily frequency during the influenza season limits the likelihood of identifying an underlying relationship. It might be more fruitful to examine an extreme cold, dry subset of Dry Polar air masses to identify the coldest and driest days.

Conclusions

In New York City, high P&I mortality periods within a given year were preceded by multiple day periods with unusually low temperature and humidity. Over the 28-year period of this study, we identified 12 episodes of high P&I mortality and found that both the total mortality occurring during each episode and duration of each episode were inversely correlated with the average morning dew point temperature prior to and during the episodes. These results support the burgeoning hypothesis that unusually cold dry air enhances the airborne transmission of influenza virus.

The exploratory nature of this analysis was necessitated by the lack of an underlying theory of influenza seasonality, socio-behavioral factors, and inherent variability in disease transmission and virulence. The time between infection and a resulting mortality event (i.e. “latency”) varies between individuals depending on age, overall health, co-morbid conditions, and other factors. Thus, lags must be estimated to best fit the overall data structure. Similarly, the high frequency variability in the variables requires some smoothing to elucidate relationships, and the selection of appropriate smoothers is somewhat subjective. Nevertheless, our findings are consistent with several others. For example, there is evidence supporting a two-week lag between rising influenza virus and pneumonia mortality [28]. Other research showed fairly convincingly the existence of a 2–4 week lag between laboratory-confirmed cases of influenza and increased incidence of invasive pneumococcal disease [43]. The weather variables used in this study do not directly account for the ambient conditions experienced by the influenza victims while indoors, but cold and/or stormy weather could result in the decedents spending more time in heated indoor environs with low humidity, thereby enhancing infection opportunities [8].

For this study, P&I mortality was used to characterize the influenza time series in New York City. The limitation of this method is the potential for confounding as P&I mortality includes mortality from infections other than influenza. In addition, in non-pandemic years, P&I mortality is skewed by the extremes of age. This limitation is unlikely to be a major contributor in this study as 90% of influenza-related deaths involve persons over the age of 65 during seasonal epidemics [45] and our mortality data are age-adjusted to account for demographic changes in New York City over the period of the study.

We chose to focus on New York City because the large population (and thus large daily P&I mortality rate) enhances statistical robustness, and New York weather is highly variable owing to its midlatitude, coastal location. These results should be confirmed using a similar methodology in other cities worldwide to determine if the humidity-influenza linkage is pervasive. It would be particularly interesting to determine how these relationships evolve in subtropical or tropical climates where the P&I mortality seasonality is more muted or nonexistent.

It is likely that the underlying causes of influenza seasonality are multi-factorial, and we suspect that weather is but one of those factors. A predictive model for P&I mortality based on weather alone would likely be unsuccessful in accounting for most of the short-term influenza variability. Nevertheless, our results confirm recent emerging hypotheses of a relationship between cold, dry air and influenza transmission or virulence [6], [8]. Identifying periods of low dew point temperatures a few days or even weeks in advance is well within the skill of existing weather forecast models. Given the lag between infection and mortality, it seems reasonable to propose that, during the prime influenza season, skillful forecasts of high P&I mortality periods could be made weeks to months in advance.

Supporting Information

Figure S1.

Sample of mortality data from December 1, 1998 through March 10, 1999. Daily mortality (z-scores) (red dashed lines) shows evidence of the beginning of a prolonged peak starting around day 43. When these data are smoothed using a 17-day centered moving average filter (solid red line), the mortality peak becomes more evident. In our analysis, we instead employ a leading 17-day smoother (blue line), which effectively shifts the red line forward by 8 days. High mortality episodes are classified when the z-scores exceeds 1, so the 1999 episodes begins on day 39 and ends on day 77.

doi:10.1371/journal.pone.0034091.s001

(TIF)

Figure S2.

Example of weather data treatment for 1998–1999. (Bottom panel) Raw dew point temperature (blue dashed line), dew point z-score (green line) and z-scored after application of a 5-day centered moving average smoother (solid black line). (Top panel) Days classified as having a moist moderate air mass present (vertical bars), moist moderate frequency anomalies to remove seasonality (red dashed line), and smoothed using a 7-day centered moving average filter (purple line). For the air mass variable, this procedure converts a nominal variable into a continuous variable for subsequent analysis.

doi:10.1371/journal.pone.0034091.s002

(TIF)

Figure S3.

Comparison of smoothed mortality (black dashed line) to smoothed dew point temperature (red line) for the 1998–1999 season. The decline in dew point on day 10 preceded the start of the mortality increase by approximately 17 days. During the subsequent period of declining dew points, mortality continued to rise. When dew point reached its minimum for this period on day 31, 17-day lagged mortality was one standard deviation above the mean. Although there is no consistent 1:1 lagged relationship between dew point temperature and mortality, this example illustrates the procedure and shows a general linkage between a low dew point period in mid-late December, 1998 and a subsequent high pneumonia and influenza mortality anomaly several weeks later.

doi:10.1371/journal.pone.0034091.s003

(TIF)

Author Contributions

Conceived and designed the experiments: RED CER KBE. Performed the experiments: CER. Analyzed the data: CER RED KBE. Wrote the paper: RED CER KBE.

References

  1. 1. Davis RE, Knappenberger PC, Michaels PJ, Novicoff WM (2004) Seasonality of climate-human mortality relationships in US cities and impacts of climate change. Clim Res 26: 61–76.
  2. 2. Dowell SF, Ho MS (2004) Seasonality of infectious diseases and severe acute respiratory syndrome–what we don't know can hurt us. Lancet Infect Dis 4: 704–708.
  3. 3. Dushoff J, Plotkin JB, Viboud C, Earn DJD, Simonsen L (2006) Mortality due to influenza in the United States—An annualized regression approach using multiple-cause mortality data. Am J Epidemiol 163: 181–187.
  4. 4. Lowen AC, Mubarka S, Steel J, Palese P (2007) Influenza virus transmission is dependent on relative humidity and temperature. Pathogens 3: 1470–1476.
  5. 5. Shaman J, Kohn M (2009) Absolute humidity modulates influenza survival, transmission, and seasonality. Proc Nat Acad Sci 106: 3645–3646.
  6. 6. Shaman J, Pitzer VE, Viboud C, Grenfell T, Lipsitch M (2010) Absolute humidity and the seasonal onset of influenza in the continental United States. PLoS Biol 8: e1000316. doi:10.1371/journal.pbio.1000316.
  7. 7. Lowen AC, Steel J, Mubareka S, Palese P (2008) High temperature (30°C) blocks aerosol but not contact transmission of influenza virus. J Virology 82: 5650–5652.
  8. 8. Lofgren E, Fefferman N, Naumov YN, Gorski J, Naumova EN (2007) Influenza seasonality: Underlying causes and modeling theories. J Virology 81: 5429–5436.
  9. 9. Fuhrmann CM (2010) The effects of weather and climate on the seasonality of influenza: What we know and what we need to know. Geogr Compass 4: 718–730.
  10. 10. Tamerius J, Nelson M, Zhou S, Viboud C, Miller MA, et al. (2011) Global Influenza seasonality: Reconciling patterns across temperate and tropical regions. Environ Health Perspect 119: 439–445.
  11. 11. Le Merre C, Kim HH, Chediak AD, Wanner A (1996) Airway blood flow responses to temperature and humidity of inhaled air. Resp Physiol 105: 235–239.
  12. 12. Eccles R (2002) An explanation for the seasonality of acute upper respiratory tract viral infections. Acta Otolaryngol 122: 183–191.
  13. 13. Assaad FA, Reid D (1971) Some factors influencing mortality from influenza. Bull Wld Hlth Org 45: 113–117.
  14. 14. Loosli C, Lemon H, Robertson O, Appel E (1943) Experimental air-borne influenza infection. I. Influence of humidity on survival of virus in air. Proc Soc Exp BioMed 53: 205–206.
  15. 15. Harper G (1961) Airborne micro-organisms: Survival tests with four viruses. J Hyg 59: 479–486.
  16. 16. Schaffer FL, Soergel ME, Straube DC (1976) Survival of airborne influenza virus: Effect of propagating host, relative humidity, and composition of spray fluids. Arch Virol 51: 263–273.
  17. 17. Hall CB (2007) The spread of influenza and other respiratory viruses: complexities and conjectures. Clin Infect Dis 45: 353–359.
  18. 18. Weber T, Stilianakis N (2008) Inactivation of influenza A viruses in the environment and modes of transmission: A critical review. J Infection 57: 361–373.
  19. 19. Polozov IV, Bezrukov L, Gawrisch K, Zimmerberg J (2008) Progressive ordering with decreasing temperature of the phospholipids of influenza virus. Nat Chem Biol 4: 248–255.
  20. 20. Ånestad G (1987) Surveillance of respiratory viral infections by rapid immunofluorescence diagnosis, with emphasis on virus interference. Epidemiol Infect 99: 523–531.
  21. 21. Dushoff J, Plotkin JB, Levin SA, Earn DJD (2004) Dynamical resonance can account for seasonalilty of influenza epidemics. Proc Natl Acad Sci 101: 16915–16916.
  22. 22. Cauchemez S, Valleron AJ, Boelle PY, Flahault A, Ferguson NM (2008) Estimating the impact of school closure on influenza transmission from Sentinel data. Nature 452: 750–754.
  23. 23. Thompson WW, Shay DK, Weintraub E, Brammer L, Cox N, et al. (2003) Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA 289: 179–186.
  24. 24. Crighton EJ, Moineddin R, Mamdani M, Upshur RE (2004) Influenza and pneumonia hospitalizations in Ontario: A time-series analysis. Epidemiol Infect 132: 1167–1174.
  25. 25. Thompson WW, Shay DK, Weintraub E, Brammer L, Bridges CB, et al. (2004) Influenza-associated hospitalizations in the United States. JAMA 292: 1333–1340.
  26. 26. Serfling RE (1963) Methods for current statistical analysis of excess pneumonia-influenza deaths. Public Health Rep 78: 494–506.
  27. 27. Sprenger MJ, Mulder PG, Beyer WE, Masurel N (1991) Influenza: relation of mortality to morbidity parameters–Netherlands, 1970–1989. Int J Epidemiol 20: 1118–1124.
  28. 28. Wong CM, Chan KP, Hedley AJ, Peiris JSM (2004) Influenza-Associated Mortality in Hong Kong. Clin Infect Dis 39: 1611–1617.
  29. 29. Denoeud L, Turbelin C, Ansart S, Valleron A-J, Flahault A, et al. (2007) Predicting pneumonia and influenza mortality from morbidity data. PLoS ONE 2: e464. doi:10.1371/journal.pone.0000464.
  30. 30. Thompson M, Shay D, Chou H, Bridges C, Cheng P, et al. (2010) Estimates of Deaths Associated with Seasonal Influenza — United States, 1976–2007. MMWR 59: 1057–1063.
  31. 31. Ma J, Dushoff J, Earn DJD (2011) Age-specific mortality risk from pandemic influenza. J Theor Biol 288: 29–34.
  32. 32. Anderson RN, Rosenberg HM (1998) Age Standardization of Death Rates: Implementation of the Year 2000 Standard. National Vital Statistics Reports Vol 47, No 3. Hyattsville, MD: National Center for Health Statistics.
  33. 33. Sheridan SC (2002) The redevelopment of a weather-type classification scheme for North America. Intl J Climatol 22: 51–68.
  34. 34. Hondula DM, Davis RE (2011) Climatology of winter transition days for the contiguous USA, 1951–2007. Theor Appl Climatol 103: 27–37. DOI 10.1007/s00704-010-0278-7.
  35. 35. Sheridan SC, Dolney TJ (2003) Heat, mortality, and level of urbanization: Measuring vulnerability across Ohio, USA. Clim Res 24: 255–265.
  36. 36. Sheridan SC, Kalkstein LS (2004) A synoptic Climatological approach to separate weather- and pollution-induced impacts on human mortality. Epidemiology 15: S40.
  37. 37. Rainham DGC, Smoyer-Tomic KE, Sheridan SC, Burnett RT (2005) Synoptic weather patterns and modification of the association between air pollution and human mortality. Intl J Environ Health Res 15: 347–360.
  38. 38. Dolney TJ, Sheridan SC (2006) The relationship between extreme heat and ambulance calls for the city of Toronto, Ontario, Canada. Environ Res 101: 94–103.
  39. 39. Michelozzi P, De Sario M, Accetta G, de'Donato F, Kirchmayer U, et al. (2006) Temperature and summer mortality: geographical and temporal variations in four Italian cities. J Epidemiol Community Health 60: 417–423.
  40. 40. Morabito M, Crisci A, Grifoni D, Orlandini S, Cecchi L, et al. (2006) Winter air-mass-based synoptic climatologial approach and hospital admissions for myocardial infarction in Florence, Italy. Environ Res 102: 52–60.
  41. 41. Kalkstein LJ, Greene JS, Mills DM, Perrin AD, Samenow JP, et al. (2007) The development of analog European heat waves for U.S. cities to analyze the impacts of heat-related mortality. Bull Amer Meteorol Soc 88: 75–85.
  42. 42. Wilks DS (2006) 627 p. Statistical Methods in the Atmospheric Sciences (2nd Edn.) Academic Press, Boston/San Diego/London.
  43. 43. Grabowska K, Högberg L, Penttinen P, Svensson A, Ekdahl K (2006) Occurrence of invasive pneumococcal disease and number of excess cases due to influenza. BMC Infect Dis 6: 58.
  44. 44. Knight DB, Davis RE, Sheridan SC, Hondula DM, Sitka SJ, et al. (2008) Increasing frequencies of warm and humid air masses over the conterminous United States from 1948 to 2005. Geophys Res Lett 35: L10702. doi:10.1029/2008GL033697.
  45. 45. Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, et al. (1998) Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 178: 53–60.