Spatial Associations of Leprosy and Schistosomiasis and Potential effects of this co-endemic helminth on the transmission of leprosy in Minas Gerais, Brazil Open Access

Stephens, Jessica L (2017)

Permanent URL:


Background Brazil has the second highest prevalence of leprosy (Hansen's Disease, HD), but factors contributing to transmission remain unclear. Pilot data from Minas Gerais, Brazil suggest a potential spatial association between HD and schistosomiasis, an important helminth infection. Studies have also shown a predisposition to the more infectious multibacillary leprosy (MB) in those co-infected with helminths, supporting biological plausibility for the increase in HD transmission in areas with co-endemic helminths.

Methods An ecological study using public health surveillance and census data was conducted to investigate whether the occurrence of HD -and specifically MB disease-is associated with the presence of schistosomiasis in a community in 41 municipalities of the state of Minas Gerais, Brazil, 2011 to 2015. Multivariate logistic regression and spatial statistics (K-function, bivariate K-function, bivariate local indicator of spatial autocorrelation [LISA], Kulldorff's spatial scan) were applied to the data.

Results The average annual incidence was high for HD at 35.3 per 100,000. Schistosoma mansoni average annual incidence was 26 per 100,000. Fifteen high-high clusters of local bivariate autocorrelation for HD and schistosomiasis were identified, while 11 clusters were detected for MB and schistosomiasis. However, there was no overlap between significant most likely clusters of HD or MB disease with schistosomiasis. While living in a census tract with reported schistosomiasis was not associated with MB disease on an individual level, census level multivariate analysis found the risk of MB presence was over 1.5 time greater in tracts with reported schistosomiasis than in tracts without, adjusted for population density, household density, and household income (aOR=1.66, 95% CI 1.01, 2.71).

Conclusion This study provides a novel means to study HD transmission using spatial analysis to analyze co-occurrence of schistosomiasis which may affect HD transmission in an area with clusters of hyperendemic HD. Bivariate LISA clusters depict areas with substantial co-occurrence. Furthermore, census tract multivariate analysis indicate that the association of schistosomiasis with MB disease warrants more detailed analysis through co-infection studies. These findings not only suggest that helminth infections are associated with HD transmission, but they also can guide control programs in co-endemic areas to decrease the burden of each infection.

Table of Contents

Table of Contents



A. Abstract 11

B. Introduction. 13

C. Methods. 15

Ethical considerations. 15

Study Area. 16

Data Sources and Methods. 16

Spatial Data Analysis. 17

Logistic Regression Data Analysis. 19

D. Results. 20

Study population. 20

Individual Level Bivariate and Multivariate Analysis. 21

Individual Level Spatial Analysis. 22

Census Tract Description. 22

Census Tract Spatial Analysis. 22

Census Tract Multivariate Analysis. 24

E. Discussion. 25

Conclusions. 29

F. Tables. 30

Table 1. Descriptive Characteristics Of Incident Cases Of Hansen's Disease In Minas Gerais, 2011-2015 (N=755). 30

Table 2. Unadjusted Bivariate Association Of Study Variables For Risk Of Multibacillary Hansen's Disease In Minas Gerais, Brazil, 2011-2015. 31

Table 3. Adjusted Odds Ratios And 95% Confidence Intervals for Schistosomiasis and Covariates with Multibacillary Hansen's Disease in Minas Gerais, Brazil, 2011-2015. 32

Table 4. Kuldorff's Spatial Scan Most Likely Clusters Of HD, MD Disease And Schistosomiasis. 33

Table 5. Spatial Autocorrelation Of HD, MD Disease And Schistosomiasis. 33

Table 6. Adjusted Odds Ratios And 95% Confidence Intervals For Schistosomiasis And Covariates With All Hansen's Disease At Aggregated Census Tract Level. 34

Table 7. Adjusted Odds Ratios And 95% Confidence Intervals For Schistosomiasis And Covariates With Multibacillary Hansen's Disease At Aggregated Census Tract Level. 35

G. Figures. 36

Figure 1. Study Area: 41 Municipalities in Minas Gerais, Brazil. 36

Figure 2. Univariate Spatial Analysis: Aggregated census tract level spatial analysis for all HD cases, MB disease and Schistosomiasis, respectively (1,2,3). Raw Annual Incidence (Column A), a Spatially Empirical Bayesian approach to smoothing (B), Local indicator of Spatial Autocorrelation (LISA, C), and Kulldorff's Spatial Scan Statistics most likely clusters (D) are presented. 37

Figure 3. Bivariate Spatial Analysis: Bivariate LISA characterizing areas with a statistically significant (p<0.05) positive spatial association with the average annual incidence of HD and schistosomiasis (A) compared to MB disease and schistosomiasis (B). 38


Summary and Public Health Implications. 39

Possible Future Directions. 40


Figure S1. Data Flow Chart 42

Figure S2. Distribution of HD Cases by Municipality, 2011-2015, Full vs Geocoded Data 43

Figure S3. Distribution of Schistosomiasis Cases by Municipality, 2011-2015, Full vs Geocoded Data 44

Figure S4. Bivariate K-function (MB disease vs PB disease) 45


About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Partnering Agencies
Last modified

Primary PDF

Supplemental Files