School of Mathematics, Statistics and Computer Science, University of KwaZulu Natal, Durban, South Africa.
School of Science, Mathematics and Statistics, University of Energy and Natural Resources, Sunyani, Ghana.
Sci Rep. 2022 Jul 7;12(1):11498. doi: 10.1038/s41598-022-15561-4.
Malaria and anaemia are common diseases that affect children, particularly in Africa. Studies on the risk associated with these diseases and their synergy are scanty. This work aims to study the spatial pattern of malaria and anaemia in Nigeria and adjust for their risk factors using separate models for malaria and anaemia. This study used Bayesian spatial models within the Integrated Nested Laplace Approach (INLA) to establish the relationship between malaria and anaemia. We also adjust for risk factors of malaria and anaemia and map the estimated relative risks of these diseases to identify regions with a relatively high risk of the diseases under consideration. We used data obtained from the Nigeria malaria indicator survey (NMIS) of 2010 and 2015. The spatial variability distribution of both diseases was investigated using the convolution model, Conditional Auto-Regressive (CAR) model, generalized linear mixed model (GLMM) and generalized linear model (GLM) for each year. The convolution and generalized linear mixed models (GLMM) showed the least Deviance Information Criteria (DIC) in 2010 for malaria and anaemia, respectively. The Conditional Auto-Regressive (CAR) and convolution models had the least DIC in 2015 for malaria and anaemia, respectively. This study revealed that children in rural areas had strong and significant odds of malaria and anaemia infection [2010; malaria: AOR = 1.348, 95% CI = (1.117, 1.627), anaemia: AOR = 1.455, 95% CI = (1.201, 1.7623). 2015; malaria: AOR = 1.889, 95% CI = (1.568, 2.277), anaemia: AOR = 1.440, 95% CI = (1.205, 1.719)]. Controlling the prevalence of malaria and anaemia in Nigeria requires the identification of a child's location and proper confrontation of some socio-economic factors which may lead to the reduction of childhood malaria and anaemia infection.
疟疾和贫血是影响儿童的常见疾病，尤其是在非洲。关于这些疾病及其协同作用的风险研究很少。这项工作旨在研究尼日利亚疟疾和贫血的空间模式，并使用单独的疟疾和贫血模型来调整其风险因素。本研究使用贝叶斯空间模型，采用集成嵌套拉普拉斯逼近法（INLA）来建立疟疾和贫血之间的关系。我们还调整了疟疾和贫血的风险因素，并绘制了这些疾病的估计相对风险图，以确定考虑中的疾病风险较高的地区。我们使用了 2010 年和 2015 年尼日利亚疟疾指标调查（NMIS）获得的数据。使用卷积模型、条件自回归（CAR）模型、广义线性混合模型（GLMM）和广义线性模型（GLM）分别研究了这两种疾病的空间变异性分布。2010 年，卷积和广义线性混合模型（GLMM）分别显示出最小的偏差信息准则（DIC），而条件自回归（CAR）和卷积模型在 2015 年分别显示出最小的 DIC。这项研究表明，农村地区的儿童感染疟疾和贫血的可能性很大[2010 年；疟疾：优势比（AOR）= 1.348，95%置信区间（CI）=（1.117，1.627），贫血：AOR = 1.455，95%CI =（1.201，1.7623）。2015 年；疟疾：AOR = 1.889，95%CI =（1.568，2.277），贫血：AOR = 1.440，95%CI =（1.205，1.719）]。控制尼日利亚疟疾和贫血的流行需要确定儿童的位置，并妥善应对一些可能导致儿童疟疾和贫血感染减少的社会经济因素。
