Latitudes mediate the association between influenza activity and meteorological factors: A nationwide modelling analysis in 45 Japanese prefectures from 2000 to 2018

https://doi.org/10.1016/j.scitotenv.2019.134727Get rights and content

Highlights

  • This investigation collected 45 prefectures' influenza data in Japan from 2000 to 2018.

  • Influenza risk driven by vapour pressure was negatively correlated with latitude.

  • Low temperature and low relatively humidity were predictors for influenza risk;

  • But their risk-latitude correlations were weaker than that of absolute humidity.

  • Latitude was a mediator despite its range in Japan is only from 26°N to 43°N.

Abstract

Background

Cold and dry conditions were well-documented as a major determinant of influenza seasonality in temperate countries but the association may not be consistent when the climate in temperate areas is closer to that in sub-tropical areas. We hypothesized latitudes may mediate the association between influenza activity and meteorological factors in 45 Japanese prefectures.

Methods

We used the weekly incidence of influenza-like illness of 45 prefectures from 2000 to 2018 as a proxy for influenza activity in Japan, a temperate country lying off the east coast of Asia. A combination of generalized additive model and distributed lag nonlinear model was adopted to investigate the associations between meteorological factors (average temperature, relative humidity, total rainfall, and actual vapour pressure, a proxy for absolute humidity) and the influenza incidence. Kendall’s tau b (τ) and Spearman correlation coefficient (rs) between latitude and the adjusted relative risk (ARR) of each meteorological factor were also assessed.

Results

A higher vapour pressure was significantly associated with a lower influenza risk but the ARR strongly weakened along with a lower latitude (τ = −0.23, p-value = 0.02; rs = −0.33, p-value = 0.03). Lower temperature and lower relatively humidity were significantly associated with higher influenza risks in over 65% and around 40% of the prefectures respectively but the strength and significance of the correlations between their ARRs and latitude were weaker than that from vapour pressure.

Conclusion

Even though the range of latitudes in Japan is small (26°N–43°N), the relationships between meteorological factors and influenza activity were mediated by the latitude. Our study echoed absolute humidity played a more important role in relating influenza risk, but we on the other hand showed its effect on influenza activity could be hampered in a low-latitude temperate region, which have a warmer climate. These findings thus offer a high-resolution characterization of the role of meteorological factors on influenza seasonality.

Section snippets

Background

Climate change has raised increasing concerns over the effect of weather conditions on public health, with the transmission of infectious diseases being one of the major focuses (Goggins et al., 2013, Lam et al., 2018a, Lam et al., 2018b, Wang et al., 2018a, Wang et al., 2018b). Of different infectious diseases, influenza is a high burden respiratory disease caused by influenza viruses, which could result in severe illnesses such as pneumonia. According to the estimate from Centers for Disease

Data

This study included 45 prefectures of Japan, which are, from north to south, Hokkaido, Aomori, Akita, Iwate, Yamagata, Miyagi, Niigata, Fukushima, Ishikawa, Tochigi, Toyama, Gunma, Ibaraki, Nagano, Fukui, Gifu, Tokyo, Yamanashi, Chiba, Kanagawa, Tottori, Kyoto, Aichi, Shimane, Hyogo, Shizuoka, Okayama, Osaka, Nara, Hiroshima, Mie, Kagawa, Yamaguchi, Tokushima, Wakayama, Ehime, Kochi, Fukuoka, Saga, Nagasaki, Oita, Kumamoto, Miyazaki, Kagoshima, and Okinawa (Fig. 1). Shiga and Saitama were

Descriptive statistics

Fig. 2A–E show the spatiotemporal distributions of ILI incidence and the meteorological factors. The incidence mostly peaked in late-January to Mid-March every year regardless of prefecture. During the study period, the incidence of influenza was generally higher in Fukui, Oita, Miyazaki, and Okinawa. Strong seasonal patterns were observed for temperature and vapour pressure, as their mean values were higher in summer times. Seasonality of relative humidity and total rainfall were less obvious;

Discussion

Seasonal influenza generates serious public health impacts worldwide. In this study, we elucidate the association between meteorological factors and influenza activity using an 18-year dataset from 45 Japanese prefectures having different climatic characteristics by their latitudes. We found that lower temperature and lower absolute humidity were statistically significantly associated with a higher risk of influenza over the majority of the prefectures in Japan. Particularly, the correlation

Conclusions

Our findings showed that even though the range of latitudes in Japan is small, the relationships between meteorological factors and influenza activity were mediated by the latitude. Particularly, absolute humidity played a more important role associated with the influenza seasonality, but its effect could be hampered in a low-latitude temperate region having a warmer climate. The results offer a high-resolution characterization of the role of meteorological factors on influenza seasonality on

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The work is supported by National Natural Science Foundation of China (31871340, 71974165). We thank Dr. Ayako Sumi's team for providing the data for analysis.

Authors’ contributions

KCC, JL, and AS conceived the study. KCC, JL, and SYFL performed the analysis. LW, JL, and AS contributed to the acquired data of the study. KCC, SYFL, and MHW contributed to the results interpretation. KCC, KMJ, and SYFL drafted the paper. All authors have read and approved the final paper.

Data sharing

All data will be shared upon the acceptance of the paper.

References (56)

  • D. Bolton

    The computation of equivalent potential temperature

    Mon. Weather Rev.

    (1980)
  • B.A. Brumback et al.

    Transitional regression models, with application to environmental time series

    J. Am. Stat. Assoc.

    (2000)
  • Centers for Disease Control and Prevention (CDC). Estimated Influenza Illnesses, Medical visits, Hospitalizations, and...
  • P.K. Chan et al.

    Seasonal influenza activity in Hong Kong and its association with meteorological variations

    J. Med.Virol

    (2009)
  • F.T. Chew et al.

    Seasonal trends of viral respiratory tract infections in the tropics

    Epidemiol. Infect.

    (1998)
  • K.C. Chong et al.

    Identifying meteorological drivers for the seasonal variations of influenza infections in a subtropical city – Hong Kong

    Int. J. Environ. Res. Public Health

    (2015)
  • K.C. Chong et al.

    A statistical method utilizing information of imported cases to estimate the transmissibility for an influenza pandemic

    BMC Med. Res. Methodol.

    (2017)
  • E.R. Deyle et al.

    Global environmental drivers of influenza

    Proc. Natl. Acad. Sci. U.S.A.

    (2016)
  • A. Gasparrini

    Distributed lag linear and non-linear models in R: the package dlnm

    J. Stat. Softw.

    (2011)
  • A. Gasparrini et al.

    Distributed lag non-linear models

    Stat. Med.

    (2010)
  • W.B. Goggins et al.

    Associations between mortality and meteorological and pollutant variables during the cool season in two Asian cities with sub-tropical climates: Hong Kong and Taipei

    Environ. Health

    (2013)
  • Higashide, M. 2019. Prefectures population, area, population density. Available: https://uub.jp/rnk/p_j.html (Accessed...
  • C.M. Hurvich et al.

    Smoothing parameter selection in nonparametric regression using an improved Akaike information criterion

    J. R. Statist. Soc.: Series B

    (1998)
  • H.C. Lam et al.

    Different sensitivities to ambient temperature between first-and re-admission childhood asthma cases in Hong Kong–a time series study

    Environ. Res.

    (2018)
  • H.C. Lam et al.

    Short-term association between ambient temperature and acute myocardial infarction hospitalizations for diabetes mellitus patients: a time series study

    PLoS Med.

    (2018)
  • M. Lipsitch et al.

    Influenza seasonality: lifting the fog

    Proc. Natl. Acad. Sci. U.S.A.

    (2009)
  • E. Lofgren et al.

    Influenza seasonality: underlying causes and modeling theories

    J. Virol.

    (2007)
  • T.P. Loh et al.

    Correlations between clinical illness, respiratory virus infections and climate factors in a tropical paediatric population

    Epidemiol. Infect.

    (2011)
  • Cited by (37)

    • Subtypes specified environmental dependence of seasonal influenza virus

      2022, Science of the Total Environment
      Citation Excerpt :

      This can explain why previous studies have found that the important drivers of influenza transmission varied among regions (Tamerius et al., 2011; Tamerius et al., 2013; Metcalf et al., 1860). The narrow meteorological space exists in most location-specified analyses (Chong et al., 2020; Almeida et al., 2018), and whether the location-specific association can be extrapolated to regions where the environmental condition is different needs to be carefully investigated. In the present studies, leveraging the various dynamical patterns of influenza and diverse meteorological spaces from 30 provinces across China, the environment-influenza relationship was pooled with sufficient coverage (Fig. S4).

    • Association between multiple meteorological variables and seasonal influenza A and B virus transmission in Macau

      2022, Heliyon
      Citation Excerpt :

      Humidity, wind velocity (WV) and atmospheric pressure (atm) showed a low correlation with the peaks of influenza activity (Ianevski et al., 2019). A study conducted in Japan found that at the 10th and 25th percentile of temp, the risk of influenza increased in over 65% of prefectures compared with the median value of temp (Chong et al., 2020a,b). At the 10th and 25th percentile of RH, around 40% of the prefectures' influenza risks were increased when compared with the median value of RH (Chong et al., 2020).

    View all citing articles on Scopus
    View full text