Research ArticleEPIDEMIOLOGY

Snakebites are associated with poverty, weather fluctuations, and El Niño

See allHide authors and affiliations

Science Advances  11 Sep 2015:
Vol. 1, no. 8, e1500249
DOI: 10.1126/sciadv.1500249
  • Fig. 1 Snakes and snakebites in CR.

    (A) The terciopelo B. asper. (B) Average annual snakebite incidence, by canton, from 2005 to 2013. County color indicates snakebite incidence rate, county boundary color indicates relative risk, and a marking described in the map legend indicates the primary cluster.

  • Fig. 2 Variables spatially associated with snakebites in CR.

    (A) Altitude. (B) Rainfall. (C) Poverty gap index. (D) Destitute housing. Coefficients are shown (in the legend of each panel) only when pseudo t values are significant (P < 0.05).

  • Fig. 3 Temporal snakebite incidence patterns in CR.

    (A) Monthly time series from 2005 to 2013; colors indicate the phases of ENSO as explained in the legend (inset). (B) Seasonality in snakebite incidence; monthly box plot shows the log-transformed number of snakebites, and colors indicate the different phases of ENSO. (C) ACF of monthly snakebites. (D) CCF between snakebites and temperature. (E) CCF between snakebites and rainfall. (C to E) Dashed lines represent the 95% confidence interval for correlations expected to arise randomly. (F) Cross-wavelet coherence analysis of snakebites and ENSO. We used SST4 as an ENSO index. In the analysis, a 6-month smoothing window was used. The cross-wavelet coherence scale is from 0 (blue) to 1 (red). Red regions in the plots indicate frequencies and times for which the two series share variability (or power). The cone of influence (in which results are not influenced by the edges of data) and the significantly coherent time-frequency regions (P < 0.05) are indicated by solid black lines.

  • Table 1 Parameter estimates for the best model explaining the monthly snakebite incidence in Costa Rica.
    ParameterEstimate ± SE
    Embedded ImageAverage logarithm of monthly snakebites1.46 ± 0.06
    Embedded ImageAutoregressive component0.24 ± 0.09
    Embedded ImageSeasonal autoregressive with a 7-month lag0.32 ± 0.10
    Embedded ImageTemperature with 1-month lag0.21 ± 0.07
    Embedded ImageRainfall with 11-month lag−0.0014 ± 0.0004
    Embedded ImageSD of the error0.44

Supplementary Materials

  • Supplementary materials for this article are available at http://advances.sciencemag.org/cgi/content/full/1/8/e1500249/DC1.

    Fig. S1. Rural and urban population trends in CR from 2000 to 2013.

    Fig. S2. Assumptions of the GWR model.

    Fig. S3. Annual snakebite incidence rate in CR, by canton, from 2005 to 2013.

    Fig. S4. Snakebite incidence hot spots in CR from 2005 to 2013; annual clusters have red boundaries.

    Table S1. Summary statistics for the GWR model (the global model indicates a linear regression model without considering spatial heterogeneity).

    Table S2. Time series model selection.

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. Rural and urban population trends in CR from 2000 to 2013.
    • Fig. S2. Assumptions of the GWR model.
    • Fig. S3. Annual snakebite incidence rate in CR, by canton, from 2005 to 2013.
    • Fig. S4. Snakebite incidence hot spots in CR from 2005 to 2013; annual clusters have red boundaries.
    • Table S1. Summary statistics for the GWR model (the global model indicates a linear regression model without considering spatial heterogeneity).
    • Table S2. Time series model selection.

    Download PDF

    Files in this Data Supplement:

Navigate This Article