Table 5 Main finding between malaria incidence and climatic variables
Item | Year of publication | Study period | Country of study | Primary author |
|---|---|---|---|---|
1 | 2002 | 1956–1998 | Colombia, Ecuador, French Guiana, Guyana, Peru, Suriname and Venezuela | Gagnon [15] |
Key results •A significant positive relationship was found between El Niño and malaria epidemics in Colombia, Guyana, Peru, and Venezuela •Floods generate malaria epidemics in the coastal region of Peru and droughts favor the development of epidemics in Colombia and Guyana. In contrast in Venezuela, malaria epidemics are delayed by drought for 1 year •In Brazil, French Guiana, and Ecuador, non-climatic factors such as fumigation, variation in drug availability, and population migration are likely to play a larger role in malaria where they did not detect an ENSO/malaria signal | ||||
2 | 2009 | 1960–2006 | Colombia | Mantilla [20] |
Key results •The negative binomial regression model (NBRM) results showed a positive association between ENSO and the total number of malaria cases in Colombia. A 1 °C change in ENSO measuring variables (ENSO_Avg or ENSO_Dom) resulted in a 17.7% or 9.3% change in expected malaria cases, respectively •Regarding the five regions analysed in the study, the Pacific Region (R1) and the Atlantic Region (R2), showed that malaria cases are positively and significantly associated with the behaviour of ENSO (high during El Niño) for both the ENSO_Avg variables as ENSO_Dom and that a 1 °C change in these variables results in a change of 22.9% and 9.6% in expected malaria cases in R1 and, 23.4% and 19.4% in R2, respectively. In contrast, no significant relationship with malaria cases was evident during the study period in the Andean, Amazonian, and Orinoco regions (R3, R4, R5) | ||||
3 | 2011 | 2003–2006 | French Guiana | Girod [16] |
Key results •Camopi: Malaria incidence showed seasonality, with a peak in January–February (i.e., at the end of the short rainy season) and another in July–August (at the end of the long rainy season) • Correlation between human bite rates (HBR) by An. darlingi and precipitation 0.46 (p < 0.01) • Apatou: Malaria incidence showed two seasonal peaks, these occurred in January–February and July–August. There was no evidence between Correlation (HBR) and precipitation • Régina: In this health centre, no clear seasonal variation was observed in the incidence of malaria during the study period and no correlation was evident between (HBR) and precipitation | ||||
4 | 2011 | 2001–2009 | French Guiana | Stefani [37] |
Key results • When studying the intraannual variability, two peaks were observed for P. vivax (January and June). While only one peak was observed for P. falciparum (January) • Meteorological and hydrological characteristics are positively correlated with the incidence of malaria • Seasonality in Camopi was determined by rainfall, the incidence of malaria was significantly higher during the rainy period where greater precipitation and lower average minimum temperature are observed (December to June) than during the dry period (July to November) (p < 0.001) • Regarding An. darlingi, a significant correlation was observed between the incidence of malaria and the human bite rate (HBR) recorded a month before (p = 0.03), observing a peak in the month of May | ||||
5 | 2011 | 2002–2007 | French Guiana | Basurko [35] |
Key results •The study highlights a clear seasonal pattern in malaria incidence in Cacao, influenced by rainfall, river levels, and temperature variations, with significant correlations found between malaria cases and specific meteorological factors in the Cacao village between 2002–2007 •The highest malaria incidence was observed in 2005 and 2006 •Univariate ARIMA Model: Significant factors include mean minimum temperatures at time t and t-12 months and mean maximum temperatures at t-1, t-2, and t-9 months •Multivariate ARIMA Model: Incidence of malaria was inversely correlated with the minimum temperature at time t, and with maximum temperature at t-2 and t-9 months •The observed data suggests a complex relationship where both very high and low temperatures can impact malaria transmission dynamics differently •The temperature variations significantly influence malaria incidence, with lower minimum temperatures being positively associated with higher malaria cases and higher maximum temperatures inversely related to malaria incidence •Precipitation: •Positive Correlation: Higher precipitation levels create standing water, providing breeding sites for mosquitoes •Seasonality: Malaria incidence often peaks during or after rainy seasons | ||||
6 | 2012 | 1990–2000 | Venezuela | Delgado [36] |
Key results •The moderate ENSO phases significantly influence malaria incidence, varying by municipality and year •The Cajigal municipality had the highest malaria incidence regardless of ENSO phase or year. However, Sucre municipality has shown high, variable malaria incidence tied to ENSO events •Regarding temperature variables, the spatial maps (1990–2000) show higher malaria cases during cold phases across municipalities. In contrast, the highest cases in 1999 were during La Niña (4,800), followed by El Niño (3,700), and neutral phases (2,000) •Sucre and Cajigal municipalities have the highest malaria during moderate ENSO events, and fewer during weak or strong events, while malaria in Sucre municipality varies with ENSO, while Cajigal remains endemic with up to 2,000 cases, unaffected by ENSO | ||||
7 | 2013 | 2003–2009 | Brazil | Filizola [29] |
Key results •It was observed that the distribution of malaria was heterogeneous in the four municipalities analysed (Coari, Codajás, Manacapuru, and Manaus) during the years studied •The incidence of malaria was influenced by the years in which extreme ENSO events (El Niño and La Niña) occurred, with 2003, 2005, and 2007 being the ones that had the highest number of malaria cases, while 2008 and 2009 showed a decrease •A significant correlation was observed between malaria incidence and temperature, especially during years with climatic extremes (2003, 2005, and 2009), because temperature increases are associated with an increase in mosquito abundance •Precipitation showed a strong correlation with malaria, being the best descriptor of malaria seasonality. It was observed that the greatest malaria transmission occurred from June to September (dry season), associated with the period after the rain •A marked relationship was also observed between water levels and malaria; this may be because the increase in water levels generates breeding sites for mosquitoes | ||||
8 | 2014 | 1990–2005 | Colombia and Etiopia | Siraj [18] |
Key results •For Colombia, only data from P. falciparum in Antioquia •Most cases were concentrated at altitudes of 1200 to 1300 m above sea level, with an average temperature of 17.6 to 18ºC •In Colombia and Ethiopia, changes in the altitudinal distribution of malaria cases were reported with average temperature over the years, a shift of the cumulative curve to the right was observed, indicating that more cases of malaria occur at altitudes. highest in a given year. This does not mean that the number of cases increased from 1994 to 1997, but rather that the distribution of the disease has moved toward a higher elevation •Scatterplots of mean altitude versus these temperatures demonstrate a movement of the distribution to higher altitudes in warmer years for the two mountain regions •The best statistical model showed a significant positive effect of mean temperature on the logarithm of malaria cases for both regions. In Colombia, this rate ranged between approximately 10 and 80% for every 1 °C increase in temperature from the highest to the lowest municipalities | ||||
9 | 2014 | 1980–2013 | Panama | Hurtado [21] |
Key results •95% of cases were caused by P. vivax. An. albimanus is the main vector in Panama •There were great differences in the seasonality of malaria during the two periods. From 1980 to 2002, epidemics were most common during December, January, and February; no differences were observed between the months. On the contrary, from 2003 to 2013 a clear seasonality was observed, with a peak of cases in February and a significant increase in the number of cases at the end of the wet season in November, the dry season (December-March) and the beginning of the rainy season in April •After a model selection process, malaria incidence for 1980–2002 was found to be a second-order autoregressive process and was also significantly associated with the El Niño 4 (SST4) index •The SST4 index was associated with interannual cycles of malaria for four-year periods ranging from 1980 to 1995 and eight-year periods between 1995 and 2006. In 1995, there was a correlation between four-year cycles in SST4 and rainfall, and a similar pattern was also observed for the Maximum Temperature. Additionally, Maximum and Minimum Temperatures were seasonally associated with SST4 for a period of one year | ||||
10 | 2015 | 2014–2015 | Brazil | Bauch [23] |
Key results •Conservation scenarios based on the estimated regression results suggest that the incidence of malaria, ARI, and diarrhea would be reduced by expanding strict protected areas, and malaria could be further reduced by restricting roads and mining • Strict protected areas (PAs) were negatively correlated with the three major diseases: malaria, diarrhea, and acute respiratory infection (ARI). This may have been due to the combined effects of reduced deforestation and exposure, meaning that strict PAs may serve as a barrier against disease. Sustainable use PAs, which allow human use and/or occupation, were positively correlated with malaria • Bioclimatic factors and natural water bodies were positively correlated with malaria. In contrast, altitude, higher temperature, and precipitation were negatively correlated with malaria • The author does not specify the species of Plasmodium causing the malaria cases, nor did he take into account entomological data | ||||
11 | 2018 | 1998–2016 | Panama | Amarilis [38] |
Key results •The authors did not observe a clear seasonality in the distribution of malaria cases, however, the maximum points reached were observed in September and March Regarding climatic variables: precipitation showed strong seasonality with dry months (December–April) and a peak in July. A positive correlation with malaria was observed with a lag of 7 months. The temperature peaks occurred in April and no significant correlation with malaria was observed. An increase in malaria cases was evident during the warm phases of ENSO when the SST4 index reached its peak | ||||
12 | 2018 | 2014–2015 | Brazil | Coutinho [32] |
Key results •A strong trend was observed in malaria cases and seasonality is evident in river levels, precipitation, and temperature in the studied areas •PRQQ Area (Includes Lake Paraquequara, which is on the edge of the city limits): Showed a highly significant model (p < 2.2e-16), where the river level is the most significant factor and explains about 50% of the occurrence of the cases, additionally, moderate correlations were also observed for air temperature and precipitation •BARC Area (Barcelos): Moderate correlations were also found with river level, while air temperature had a weak correlation •In the SGC Area (Sao Gabriel da Cahoeria): No significant statistical relationships were found. However, he observed an inverse effect between malaria cases and river levels and a positive correlation with precipitation and air temperature •The river level variable had an immediate positive correlation with malaria incidence, while precipitation and temperature showed delayed effects •Temperature: Higher temperatures (> 28.5 °C) were observed to be associated with a reduction in malaria cases in BARC and PRQQ, but not in SGC •Precipitation: It was observed that PRQQ recorded the greatest increase in cases from December to January when there was constant rainfall and a slight drop in temperature and the river level went from dry to flooded | ||||
13 | 2019 | 2010–2017 | Peru | Solano [26] |
Key results •For the study period 2010–2017, 321,210 cases of malaria were reported in 2,766 (96.9%) georeferenced villages of Loreto •Cases increased from 10,994 in 2011 to 59,257 in 2014 and 58,679 in 2015, then in 2017 they decreased slightly to 51,663 •CAR (Cummulative annual rainfall) was the highest predictor, ranging from 17% to 48.4% for P. vivax and from 11.5% to 30.7% for P. falciparum •The highest risk areas identified for malaria using BRT models were: Zone I (Maynas): 42.9% high risk for P. vivax and 11.7% high risk for P. falciparum Zone II (Loreto): 56.5% high risk for P. vivax, and 27.8% high risk for P. falciparum Zone III (Datem del Marañón and Alto Amazonas): 34.5% high risk for P. vivax and 5.4% high risk for P. falciparum Zone IV (Requena and Ucayali): 3.9% high risk for P. vivax Zone V (Ramón Castilla): 45.3% high risk for P. vivax, a significant number of towns at risk | ||||
14 | 2020 | 2006–2018 | Ecuador | Gunderson [27] |
Key results •Between 2006–2018, 9,230 cases of P. vivax malaria and 499 cases of P. falciparum were reported in Ecuador •P. vivax: The incidence increased in 2008, decreased at the end of 2011, and increased again in 2014 •P. falciparum: maintained low levels until a peak in early 2016, then remained low •It was observed that the cantons bordering Loreto, Peru, had a higher incidence of malaria than the non-border cantons (3.1% increase for P. vivax, 3.0% for P. falciparum per increase of 1 case/1000/ month) •The Aguarico canton presented the highest rates, with peaks of 7.4 cases/1000/week (P. vivax) and 2.3 cases/1000/week (P. falciparum) •Precipitation had a stronger effect on P. falciparum, while higher soil temperatures decreased the incidence of P. vivax, similar effects on soil moisture were observed for both species | ||||