Improving socioeconomic status may reduce the burden of malaria in sub Saharan Africa: A systematic review and meta-analysis

Abstract

Background: A clear understanding of the effects of housing structure, education, occupation, income, and wealth on malaria can help to better design socioeconomic interventions to control the disease. This literature review summarizes the relationship of housing structure, educational level, occupation, income, and wealth with the epidemiology of malaria in sub-Saharan Africa (SSA).

Methods: A systematic review and meta-analysis was conducted following the preferred reporting items for systematic reviews and meta-analyses guidelines. The protocol for this study is registered in PROSPERO (ID=CRD42017056070), an international database of prospectively registered systematic reviews. On January 16, 2016, available literature was searched in PubMed, Embase, CINAHL, and Cochrane Library. All but case studies, which reported prevalence or incidence of Plasmodium infection stratified by socioeconomic status among individuals living in SSA, were included without any limits. Odds Ratio (OR) and Relative Risk (RR), together with 95% CI and p-values were used as effect measures. Heterogeneity was assessed using chi-square, Moran's I2, and tau2 tests. Fixed (I2<30%), random (I2≥30%) or log-linear dose-response model was used to estimate the summary OR or RR.

Results: After removing duplicates and screening of titles, abstracts, and full text, 84 articles were found eligible for systematic review, and 75 of them were included in the meta-analyses. Fifty-seven studies were cross-sectional, 12 were prospective cohort, 10 were case-control, and five were randomized control trials. The odds of Plasmodium infection increased among individuals who were living in poor quality houses (OR 2.13, 95% CI 1.56-3.23, I2 = 27.7), were uneducated (OR 1.36, 95% CI 1.19-1.54, I2 = 72.4.0%), and were farmers by occupation (OR 1.48, 95% CI 1.11-1.85, I2 = 0.0%) [p<0.01 for all]. The odds of Plasmodium infection also increased with a decrease in the income (OR 1.02, 95% CI 1.01-1.03, tau2<0.001), and wealth index of individuals (OR 1.25, 95% CI 1.18-1.35, tau2 = 0.028) [p<0.001 for both]. Longitudinal studies also showed an increased risk of Plasmodium infection among individuals who were living in poor quality houses (RR 1.86, 95% CI 1.47-2.25, I2 = 0.0%), were uneducated (OR 1.27, 1.03-1.50, I2 = 0.0%), and were farmers (OR 1.36, 1.18-1.58) [p<0.01 for all].

Conclusions: Lack of education, low income, low wealth, living in poorly constructed houses, and having an occupation in farming may increase risk of Plasmodium infection among people in SSA. Public policy measures that can reduce inequity in health coverage, as well as improve economic and educational opportunities for the poor, will help in reducing the burden of malaria in SSA.

Fig 1. PRISMA flow diagram showing the number of articles retrieved, screened, excluded, and included at each stage of the search of published articles examining the relationship of socioeconomic status with the epidemiology of malaria in sub-Saharan Africa.

Fig 2. Forest plot showing the relationship of the roof and walls nature of a house with the epidemiology of malaria in sub-Saharan Africa.

Values show the odds ratio of Plasmodium infection (95% CI). Subtotal (summary) ORs estimated using random effect models. Weights estimated using inverse variance method. I², a measure of heterogeneity.

Fig 3. Forest plot showing the relationship of windows, floor, ceiling and eaves nature of a house with the epidemiology of malaria in sub-Saharan Africa.

Values show the odds ratio of Plasmodium infection (95% CI). Subtotal (summary) ORs estimated using random effect models when I² ≥30 and using fixed effect models when I² <30. Weights estimated using inverse variance method. I², a measure of heterogeneity.

Fig 4. Forest plot comparing the odds ratio of Plasmodium infection between individuals without formal education or illiterate with those who had primary and secondary or more education level.

Subtotal (summary) ORs estimated using random effect model. Weights estimated using inverse variance method. I², a measure of heterogeneity.

Fig 5. Forest plot comparing the odds ratio of Plasmodium infection between individuals with primary education level versus those with secondary or more, and those with secondary versus tertiary or more education level Subtotal (summary) ORs estimated using random effect models when I² ≥30 and using fixed effect models when I²<30.

Weights estimated using inverse variance method. I², a measure of heterogeneity.

Fig 6. Forest plot showing the relationship of wealth and occupation with the epidemiology of malaria in sub-Saharan Africa.

Wealth was treated as continuous and occupation was treated as a categorical variable. Values show the odds ratio of Plasmodium infection (95% CI). Total (summary) ORs estimated using random effect model. Weights estimated using inverse variance method. I-Squared, a measure of heterogeneity.