INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,

MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue IX, September 2024

www.ijltemas.in Page 94

Variability Effects of Meteorological Factors on Reported Malaria

Cases in Kano and Lagos States of Nigeria

Adetunji K. Ilori

, Ugochukwu D. Uche

, Adebisi Michael

, Damilare M. Oladimeji

, Omaku P. Enesi

, Toyosi Adebambo

Statistics Programme, National Mathematical Center, Abuja, Nigeria Kaduna-Lokoja Expressway, Sheda, Kwali, Abuja.

Mathematics Programme, National Mathematical Center, Abuja, Nigeria Kaduna-Lokoja Expressway, Sheda, Kwali, Abuja.

Nigeria Centre for Disease Control and Prevention

Department of Statistics, University of Abuja, FCT, Nigeria

Department of Mathematics and Statistics, Federal Polytechnic, Nassarawa

UNICAF University, Zambia

DOI: https://doi.org/10.51583/IJLTEMAS.2024.130909

Received: 24 August 2024; Accepted: 10 September 2024; Published: 05 October 2024

Abstract: Malaria is one of life’s threatening disease that contributed significantly to morbidity and mortality in Nigeria. The variability

effects of meteorological factors between the Northern and Southern region of Nigeria on malaria transmission remain unclear and not

well researched. This research was aimed at comparing the non-linear effects of seasonal, trend and meteorological variables on monthly

malaria cases in Kano and Lagos State of Nigeria. The monthly malaria cases for age group above 5-year for Kano and Lagos State were

obtained from DHIS for the period of January, 2016 to December, 2022. Also, the predictors/meteorological variables (Rainfall, Relative

Humidity, Minimum Temperature and Maximum Temperature) were obtained from Nigeria Meteorological Agent. Generalized Additive

Model was fitted using Negative Binomial to apply smooth function to estimate and compare the effects of the selected predictors on the

malaria transmission in Kano and Lagos States. Rainfall and Maximum temperature were found not significant in explaining the effect

of malaria transmission. Malaria transmission was found to be seasonal and varying in Kano State with highest transmission between

September and October while the seasonal effect was reduced to linear in Lagos State. The long term trend had a significant increasing

effect on malaria transmission in Kano Sate compared to non-significant effect in Lagos State. The effect of Relative Humidity was

reduced to linear and increasing in the both State. Paying attention to the variability in the influence of season, trend and meteorological

variables on malaria transmission will enhance respective State Government to develop an early warning and awareness that will help in

controlling the malaria outbreak in their State.

Keywords: Malaria Transmission, Meteorological Factors, Negative Binomial Model, Generalized Additive Model, Repeated Measures

Correlation

I. Introduction

Malaria remains a significant public health challenge in Nigeria, with the country accounting for a staggering 25% of global malaria

cases and 30% of malaria-related deaths [10]. The transmission of this debilitating disease is influenced by various environmental and

ecological factors, with meteorological variables playing a crucial role [3]. Temperature, rainfall, and humidity have been identified as

key determinants that affect the survival, development, and distribution of the Anopheles mosquito vector, as well as the parasite's

reproduction and transmission dynamics [7]. Ultimately, the significance of this research lies in its potential to contribute to the ongoing

fight against malaria in Nigeria, a country that bears a disproportionate burden of the disease. By providing a deeper understanding of the

meteorological drivers of malaria transmission in Kano and Lagos States, this study can pave the way for more effective and sustainable

malaria control strategies, ultimately improving public health outcomes and contributing to the achievement of the Sustainable

Development Goals related to combating malaria and other neglected tropical diseases.

Research in Nigeria has consistently demonstrated a seasonal variation in malaria cases, with studies highlighting the impact of

meteorological factors on disease transmission. [8] found that the highest parasite densities occurred in September, while [6] reported that

mosquito densities and malaria transmission risks peaked during the rainy season. [9] also noted a surge in asymptomatic parasitaemia in

July, corresponding to the wet season. Furthermore, there have been emphases on the role of seasonal variations in malaria infection, with

recommendations targeted interventions in affected states. This studies collectively underscore the need for season-specific malaria

control measures in Nigeria. [1] highlighted, rainfall as a chief factor influencing the transmission of malaria, as it provides breeding sites

for mosquito vectors. Temperature, on the other hand, determines the duration of development of mosquito larvae in the environment,

while precipitation, directly related to rainfall, is an essential factor influencing the bionomics of malaria-transmitting mosquito vectors.

Despite the evidence of the impact of meteorological variables on malaria transmission in Nigeria, limited research has been conducted

on the specific effects of these factors along the climatic variation like the Sahelian Climate prevalent in the North-East and North-West

regions Kano state inclusive, and the Tropical Rainforest Climate of the South-South and South-East regions covering the coastal states

including Lagos state, which have distinct climatic conditions and malaria burdens [5]. This study aims to bridge this gap by investigating

the relationship between meteorological variables (temperature, rainfall, and humidity) and malaria transmission along these climatic

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,

MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue IX, September 2024

www.ijltemas.in Page 95

variations within Nigeria. Understanding these relationships can inform the development of targeted interventions and early warning

systems for malaria control in these regions, ultimately contributing to the ongoing efforts to combat this persistent public health

challenge.

Significance of the Study

Malaria remains a significant public health burden in Nigeria, with substantial socioeconomic and human costs [10]. Despite progress in

malaria control efforts, the disease continues to pose a significant challenge, particularly in regions with diverse climatic conditions and

varying transmission dynamics. Kano and Lagos States, two distinct regions in Nigeria, have different malaria burdens and climatic

patterns, necessitating tailored interventions and control strategies [5].

The motivation behind this study lies in the need to understand the in the variability in the effect and relationship between meteorological

variables and malaria transmission in these two states. Previous research has established the influence of factors such as temperature,

rainfall, and humidity on the life cycle and behavior of the Anopheles mosquito vector, as well as the development and transmission of

the malaria parasite [3]. However, the effects of these variables can vary depending on the local environmental conditions and climatic

patterns.

By investigating the impact of meteorological variables on malaria transmission in Kano and Lagos States, this study aims to provide

valuable insights that can inform the development of targeted and effective malaria transmission awareness and control strategies.

Identifying the specific meteorological drivers of malaria transmission in these regions and how the effect of these drivers varies between

the regions can aid in the planning of early public warning systems, enabling their respective State Government and authorities in public

health to anticipate and prepare for potential malaria outbreaks.

II. Methodology

This study considered a retrospective panel data analysis using secondary data of monthly malaria cases in Kano and Lagos States of

Nigeria for the period of January 2016 to December 2022 among age group above 5-year. The monthly malaria incidence for the

considered states were retrieved from DHIS. The monthly climate data for Lagos, Kano and Rivers State were also derived from the

relevant government agency in Nigeria for the period of January 2016 to December 2022. The considered meteorological variables were

Rainfall, Relative Humidity (RH), Minimum Temperature and Maximum Temperature.

Generalized Additive Model

Let consider 𝑋

𝑖1

to be log-transformation and 𝑋

𝑖2

to be a reciprocal transformation in a linear model,

Then the resulting model is given as

0 1 1

log( )

i i i

Y X e





   

(1)

0 1 1 2 2

( ) ( )

i i i i

Y g X g X e

  

   

(2)

Where

()gX



and

()gX



are two functions specified by



and



The transformation for the jth predictor is given as

( ) ( )

j jk jk

g X X







(3)

Where

(.)









are B-spline basis functions,

Let

 

 



be considered as simple regression data

Then cubic spline smoother

()gx

that minimizes the above is given as

 

'' 2

( ) ( )

y g x g x dx

















(4)

And good fit is achieved by minimizing the sum of squares

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,

MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue IX, September 2024

www.ijltemas.in Page 96

 

()

y g x







(5)

Where

'' 2

()g x dx







measures how wiggly

()gx

And





is how much we will penalize

()gx

for being wiggly

The monthly malaria incidence is considered as the response variable and meteorological factors as the explanatory variables. The effect

of seasonal, trend and meteorological variables on monthly malaria cases were assessed using Negative Binomial Generalized Additive

Model. The predictor variables effect on malaria cases was compared between two states in Nigeria (Kano and Lagos). Lagos State is

considered to represent Southern region and Kano for Northern region.

Poisson, Negative Binomial, quasi-Poisson distributions among others are different distributions suitable for modelling count response

variable. However, negative binomial is more suitable for over-dispersed data [11].[4] shows that Negative Binomial provided a better fit

to modelling monthly malaria incidences with respect to climate variables.

The expected malaria cases count is specify as follows:

 

()

ij i i j

g E Y I State i f predictor I State i





    



(6)

Where, g is the link function,

 

()

is the expected value of malaria cases,



is the overall intercept,



are the coefficients of the

State indicator variables,

 

f predictors

are the smooth function of the selected predictors specific to each

State i

and

 

I State i

are indicator functions for each State, which are 1 if the observation belongs to State

and 0 otherwise.

For more comprehensive review on Generalized Additive Model for monthly malaria incidence, please refer to [2] and [12]

III. Results and Discussion

Overall, 10,515,037 malaria confirmed cases among age category five years and above were reported in all considered states in the study,

in Nigeria between 2016 to 2022 with highest number of cases from Kano State (6,716,089) and Lagos State (3,798,948). Table 1

revealed that high malaria transmission occurred in the state that represent northern sub region, compared to Lagos State that represent

South sub region in the study.

Table 2 contains the repeated measures correlation coefficients (r) and their significant values (p-values) for malarial cases and

meteorological variables. Malaria cases were found to have significant correlation with Rainfall, Relative humidity (RH) and Maximum

Temperature, though the strength of their correlation were seen to be from weak to moderate. Repeated correlation coefficient revealed a

weak and non-significant correlation between malaria cases and Minimum Temperature (r = -0.1159 and p-value = 0.1357). The table

further revealed that there will be no fear of multi-collinearity in including all the meteorological variables in the model as there is no

perfect correlation within the variables.

The selection of variables that will be considered for comparing the variation in the trend of malaria incidence between Lagos and Kano

was done by setting “select = TRUE” in the gam function of R package. Table 4 revealed the contributions and significant of

meteorological variables in explaining malaria incidence in the region. Season, Long term trend, RH and Minimum temperature were

found to have smooth relationship with malaria incidence, since their p-values were less than 0.05 while rainfall and maximum

temperature were found non-significant in explaining the effect of malaria. 79% of variation in malaria case was predictable using model

that consist of main effect of seasonal, trend, RH, Rainfall, Maximum temperature, Minimum temperature variables and 73.3%

percentage deviance explained while the model that exclude Rainfall and Maximum temperature accounted for the same variation and

percentage deviance explained. Furthermore, fig. 1 revealed the flat shapes of rainfall and maximum temperature on the vertical axis.

Hence, rainfall and minimum temperature will be excluded in the further analysis to compare geographical variation in malaria

transmission in Nigeria.

Table 1: Yearly Malaria cases according to States

Malaria cases among age category five and above by State

Year/State

Lagos

Kano

Total

2016

511450

500881

1012331

2017

537876

664447

1202323

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,

MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue IX, September 2024

www.ijltemas.in Page 97

2018

580744

796470

1377214

2019

596318

1013154

1609472

2020

472992

1067160

1540152

2021

539027

1295066

1834093

2022

560541

1378911

1939452

Total

3798948

6716089

10515037

Table 2. Correlation coefficients between Malaria cases and Meteorological Variables

FACTORS

MALARIA CASES

RAINFALL

MAX TEM

MIN TEMP

MALARIA CASES

1.0000

0.0001

***

0.1915

0.0313

0.4252

0.0001

***

-0.2571

.0007

-0.1159

0.1357

RAINFALL

0.1915

0.0313

1.0000

0.0001

***

0.5806

0.0001

***

-0.2417

.0016

0.0313

0.6876

0.4252

0.0001

***

0.5806

0.0001

***

1.0000

0.0001

***

-0.1791

0.0205

0.4002

0.0001

***

MAX TEMP

-0.2571

0.0007

-0.2417

.0016

-0.1791

0.0205

1.0000

.0001

***

0.5444

0.0001

***

MIN TEMP

-0.1159

0.1357

0.0313

0.6876

0.4002

.0001

***

0.5444

.0001

***

1.0000

0.0001

***

Table 3. Contributions of seasonal, trend and meteorological variables on the combined States

Factor

Estimate

P – value

Intercept

10.9503

<0.0001

Smooth Terms

edf

P - value

S(Month)

4.0529

<0.0001

S(Year)

1.6849

<0.0001

S(Rainfall)

0.0005

0.3814

S(RH)

2.2035

<0.0060

S (Max Temp)

0.0006

0.2268

S (Min Temp)

4.9721

0.0001

R-square(adj) = 0.79, Percentage of Deviance Explained = 73.3%

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,

MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue IX, September 2024

www.ijltemas.in Page 98

Figure 1: effects of seasonal, trend and meteorological variable on the combined states

State Differences in Malaria Transmission

Generalized Additive Model was fitted using Negative Binomial to apply smooth function to estimate the effects of the selected

meteorological variables (Seasonal, Trends, RH and Minimum Temperature) on the malaria cases in Kano and Lagos State.

Table 4(Month) revealed that monthly malaria cases had a significant smoothed (non-linear) relationship with season with the edf of 5 in

Kano State while malaria cases had non-significant linear relationship with season in Lagos State. The effect of season with malaria cases

was reduced to linear in Lagos State and Figure 2(Month) further revealed that it had an increasing linear effect on the Log-link scale of

the monthly malaria cases in the State while there was a sharp increase in malaria cases around the month of May and peaked between

September and October and then decreases after October in Kano State.

The effective degree of freedom from Table 4(Year) revealed that trend had a significant smoothed relationship with malaria transmission

in Kano State with higher malaria transmission from the year 2019 and above. However, the long term trend had a non-significant

smoothed relationship with malaria transmission and almost a flat linear shape on the vertical axis in Lagos State. Figure 2(RH) shows

that malaria cases had an increasing linear relationship with RH in the both states and the effects was found to be significant in Kano

State compare to Lagos State. There is higher malaria transmission in Kano State than Lagos State at RH that is above 60. Also a lower

RH between 20 to 40 was found favorable for malaria transmission in Lagos State than Kano State. There is varying and significant in the

effects of malaria cases with the minimum temperature in Kano State has shown in Table 4(Minimum Temperature). Minimum

temperature between 12 to 17 were found to neither increase or decrease the malaria transmission in the State. Minimum Temperature

above 17 to 22 had higher malaria transmission while the transmission was lower between 23 to 34 in Kano State. However, the

relationship between malaria cases and minimum temperature was reduced to a linear and decreasing in Lagos State. There is lower

malaria transmission with minimum temperature less than 25 in Lagos State.

Table 4: Smoothed effects of Seasonal, long term trend, relative humidity and minimum temperature in Kano and Lagos states

Month

Smooth Terms

Estimate

P-value

Intercept

10.7182

0.0001

Kano

0.4906

0.0001

Smooth Terms

S(Lagos)

S(Kano)

edf

1.0000

5.5640

P-value

0.1420

0.0001

Year

Smooth Terms

Estimate

P-value

Intercept

10.7194

0.0001

Kano

0.5155

0.0001

Smooth Terms

S(Lagos)

S(Kano)

edf

1.0000

2.1900

P-value

0.876

0.0001

Smooth Terms

Estimate

P-value

Intercept

10.6421

0.0001

Kano

0.7884

0.0001

Smooth Terms

S(Lagos)

S(Kano)

edf

1.0000

P-value

0.5910

0.0001

Minimum Temperature

Smooth Terms

Estimate

P-value

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,

MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue IX, September 2024

www.ijltemas.in Page 99

Intercept

10.8497

0.0001

Kano

0.3164

0.7520

Smooth Terms

S(Lagos)

S(Kano)

edf

1.0000

4.837

P-value

0.2720

0.0001

Figure 2(Month): Smoothed effects of Month on Malaria Transmission

Figure 2(Trend): Smoothed effects of Trend on Malaria Transmission

Figure 2(RH): Smoothed effects of Relative Humidity on Malaria Transmission

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,

MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue IX, September 2024

www.ijltemas.in Page 100

Figure 2 (Min Temp): Smoothed effects of Minimum Temperature on Malaria Transmission

IV. Conclusion

This research used a generalized additive model to estimate seasonal, trend, and smoothed effects of meteorological variables on malaria

transmissions in Kano and Lagos State. Kano and Lagos state were on different geographical location in Nigeria. Seasonal, trend,

Relative Humidity and Minimum Temperature were identified as factors that contribute to malaria transmissions in the two state. While

there was a sharp increase in malaria transmission around the month of May and peaked between September and October in Kano state,

Lagos state shows an increasing malaria transmission from January to December. There is high malaria transmission from year 2019 and

low malaria transmissions below year 2019 in Kano state, while there is non-significant increase or decrease in malaria transmission in

Lagos state. Relative Humidity was found to linearly increase malaria transmissions in both states. There is varying effect of minimum

temperature on malaria transmissions in Kano state. Minimum temperature between 17 to 22 had higher malaria transmissions and lower

transmissions in minimum temperature between 23 to 34 in Kano state, while the effect in Lagos state was found to be linearly

decreasing.

Conflict Of Interest

The authors declare that there is no conflict of interest

References

1. Dabaro, D., Birhanu, Z., Negash, A. (2021). Effects of rainfall, temperature and topography on malaria incidence in elimination

targeted district of Ethiopia. Malar J 20, 104 https://doi.org/10.1186/s12936-021-03641-1.

2. Carrasco-escobar, G., Qquellon, J., Villa, D., & Cava, R. (2021). Time-Varying Effects of Meteorological Variables on Malaria

Epidemiology in the Context of Interrupted Control Efforts in the Amazon Rainforest, 2000 – 2017. 8(September). frontiers in

Medicine, 1–13. https://doi.org/10.3389/fmed.2021.721515

3. Craig, M. H., et al. (2019). Climate change and malaria: a systematic review. Malaria Journal, 18(1), 1-15

4. Eunice, A., Wanjoya, A., & Luboobi, L. (2017). Statistical Modeling of Malaria Incidences in Apac District, Uganda. Scientific

Research Publishing,901–919. https://doi.org/10.4236/ojs.2017.76063

5. Measure DHS/ICF International. (2013). Demographic and Health Surveys Methodology Standard Recode Manual for DHS6

Version1.0, 145. Retrieved from https://dhsprogram.com/pubs/pdf/ DHSG4/Recode7 DHS 10Sep2018DHSG4.pdf

6. Olayemi, I. K., Ande, A. T., Chita, S., Ibemesi, G., Ayanwale, V. A., & Odeyemi, O. M. (2011). Insecticide susceptibility profile

of the principal malaria vector, anopheles gambiae s.l. (diptera: Culicidae), in north-central nigeria. Journal of Vector Borne

Diseases, 48(2), 109–112.

7. Paaijmans, K. P., et al. (2010). Temperature-dependent development of Anopheles gambiae (Diptera: Culicidae) and its impact

on malaria transmission. Journal of Medical entomology, 47(5), 937-944.

8. Samdi, L. M., Ajayi, J. A., Oguche, S., & Ayanlade, A. (2012). Seasonal variation of malaria parasite density in paediatric

population of Northeastern Nigeria. Global Journal of Health Science, 4(2), 103–109. https://doi.org/10.5539/gjhs.v4n2p103

9. Salako, L. A., Ajayi, F. O., Sowunmi, A., & Walker, O. (1990). Malaria in Nigeria: A revisit. Annals of Tropical Medicine and

Parasitology, 84(5),435–445. https://doi.org/10.1080/00034983.1990.11812493

10. WHO (2020). World Malaria Report 2020. World Health Organization

11. Xia, Y., Morrison-beedy, D., Ma, J., Feng, C., Cross, W., & Tu, X. (2012). Modeling Count Outcomes from HIV Risk Reduction

Interventions : A Comparison of Competing Statistical Models for Count Responses. Hindawi Publishing Corporation, AIDS

Research and Treatment, Volume 2012, Article ID 593569. https://doi.org/10.1155/2012/593569

12. Zerihun, T., Zewotir, T. T., & Kebede, E. (2023). Heliyon Seasonal and spatial variations of malaria transmissions in northwest

Ethiopia : Evaluating climate and environmental effects using generalized additive model. Heliyon, 9(4), e15252.

https://doi.org/10.1016/j.heliyon.2023.e15252