Social and environmental conditions related to Mycobacterium leprae infection in children and adolescents from three leprosy endemic regions of Colombia

An observational cross-sectional study was conducted during 2015 and 2016 with a sample frame made of 82 selected children and adolescents 18 years and younger. Each had HHC with a leprosy patient that had been diagnosed in the last 5 years, and all were included for convenience sampling. The study population was derived from three geographical regions where the reporting of new leprosy cases is greater than other parts of Colombia: Uraba-Antioquia-Choco (UAChR; n = 18 children), from the Caribbean Region (CR; n = 43), and from the Andean Region (AR; n = 21).

Sociodemographic characteristics were gathered using a HHC evaluation form implemented by the leprosy research team of the Instituto Colombiano de Medicina Tropical (ICMT). For the purposes of this study we focussed on: age, sex, socioeconomic status, geographic location, recorded armadillo consumption, BCG immunization status, and clinical form of the index case. The socioeconomic categories established by the Colombian government are named status 1. Low-low, 2.Low, 3. Medium-low, 4. Medium, 5. Medium-high, 6. High. Status 1, 2 and 3 corresponding to low status that included the people of scarce resources [11].

Serum samples were collected for subsequent serologic evaluation. Briefly each well of 96-well ELISA plates (Nunc-Immuno 96-well, Polysorp plates) was coated with 1 μg/ml NDO-LID antigen at room temperature then blocked using 100 μl blocking buffer (1% bovine serum albumin, BSA/ phosphate-buffered saline, PBS/ PBS + Tween20, PBS-T). Plates were incubated for 1 h with agitation at room temperature. Plates were washed (5 PBS-T + 2 PBS), and 50 μl serum (1:200 dilution in BSA 0.1%/ PBS/ PBS-T) was added, followed by a 1 h of incubation with agitation at room temperature. Subsequently, 50 μl horseradish peroxidase (HRP)-conjugated detector diluted in BSA 0.1%/ PBS/ PBS-T was added and plates incubated for 1 h with agitation at room temperature. Three detectors were evaluated: anti-human IgG, anti-human IgM and protein A (Rockland Immunochemicals Inc., Limerick, PA, USA). After incubation and washing, 50 μl TMB substrate was added for 15 min in the dark at room temperature, then stopped by adding 25 μl 1 N sulfuric acid. Optical densities (OD) were measured at 450 nm using an ELISA plate reader (Spectrophotometer Bio-Rad Xmark) [10]. Cut-off values were assessed as the average OD plus two standard deviations obtained from sera (n = 100) of healthy individuals that resided in an area not endemic for leprosy. Cut-off values of 0.127, 0.226 and 0.183 were obtained for Protein A, IgM and IgG, respectively.

Data was analyzed using Excel and SPSS 24.0 software. Univariate analysis of qualitative variables was made via absolute and relative frequencies calculation. For quantitative variables summary measures like central tendency were performed. The distribution of variables was obtained using a Kolmogorov-Smirnov test. Bivariate analysis between qualitative variables was made by applying a Pearson x² test and quantitative variables were analyzed using a Student’s t-test or Mann-Whitney U test, taking into account the distribution of these variables. Multivariate analysis was conducted using a multiple regression and binomial logistic approach. Significance level of p-value < 0.05 were established for all analysis and a risk approximation was made using an odds ratio calculation (OR) with its own confidence interval (95%CI).

This study was considered of minimal risk and was approved by the Instituto Colombiano de Medicina Tropical – Universidad CES ethics committee. After the aims of the study were explained, an informed consent form was signed by the guardians of participating children.

A total of 82 children and adolescents were included, of which 36 (44%) were male and 46 (56%) were female. The age range was 1 to 18 years of age, with a mean of 10.7 years (Table 1). Of note, the majority of these children and adolescents lived with an index leprosy case: 57 with a lepromatous case; 12 with a pure neural leprosy case, 10 with a dimorphic leprosy case, 2 with an intermediate leprosy case and 1 with a tuberculoid case.

In accordance with the socioeconomic categories established by the Colombian government, we found that 42 (51%) of the children and adolescents could be considered of low-low socioeconomic status while 40 (49%) could be considered of low or a medium-low status. When seropositivity against NDO-LID was compared in these two sample groups, the frequency of anti-NDO-LID antibodies detected by Protein A (i.e. of both IgG and IgM isotypes) and IgM was highest among children of lower socioeconomic backgrounds (Table 2;p-value =0.001).

In addition to the simple presence of antibodies, we determined anti-NDO-LID antibody levels. Significantly stronger anti-NDO-LID Protein A (p = 0.001), IgM (p = 0.011) and IgG (p = 0.032) responses were detected in children and adolescents of low-low status relative to those observed in samples from subjects of low or medium-low status (Fig. 1a). These data imply that, in addition to a greater rate of M. leprae infection, the overall levels of infection in each child or adolescent of low-low socioeconomic status are higher.

When assessing anti-NDO-LID antibodies detected by protein A and IgM according to the geographic origin of subjects, we found higher frequency of seropositivity in children and adolescents from UAChR (P = 0.0001, see Table 2).

Antibody titers were also compared based on the geographic origin of the subject. We found that anti-NDO-LID responses detected by Protein A, IgM and IgG in children and adolescents from UAChR region were significantly greater (all p-values = 0.0001) than the levels measured among samples from the Caribbean and Andean regions (Fig. 1b). Taken together, these results suggest that children and adolescents from UAChR are under higher pressure for M. leprae infection than those that reside around the Caribbean or in the Andes.

Based on observations made on socioeconomic status in each region of study, we noticed that low-low, low and medium-low status differ according to the development level of each region. We therefore decided to compare seropositivity of children and adolescents belonging to a low-low socioeconomic status for each geographic area studied (Table 3). Subsequently, comparison of seropositivity rates by low-low status for each regions revealed a higher seropositivity for IgM and protein A in the UAChR compared with the seropositivity rates found in the Caribbean and the Andes (p-value < 0.05). Significant differences in the magnitude of antibody responses of low-low status individuals were also observed across the regions. When compared to Andean and Caribbean population, significantly higher levels of anti-NDO-LID antibodies were detected in individuals belonging to low-low socioeconomic status that resided in UAChR (Fig. 1c). This data further implies that a higher M. leprae frequency of infection occurs in children and adolescents with social inequalities found in a determined geographical region.

Upon questioning, we assessed that 14 individuals reported consumption of armadillo meat at some point (17.1% of the overall cohort, although for 4 (4.9%) this data was not documented). We found that anti-NDO-LID antibodies could be detected by Protein A in 9 of 14 (64%) armadillo consumers in contrast to only 8 of 64 (12%) in those that had not eaten armadillo meat (Table 4, p-value < 0.0001). Furthermore, overall anti-NDO-LID antibody levels were higher in children and adolescents that had consumed armadillo meat compared to those that had not (p-value = 0.009 for Protein A, 0.001 for IgM and 0.013 for IgG) (Fig. 2a). These findings indicate that armadillo consumption is strongly associated with both the rate and level of anti-NDO-LID antibodies.

A recommendation within leprosy control programs has been the secondary (or initial) BCG vaccination for contacts of diagnosed index cases. Accordingly, we found that 54 (66%) of evaluated children and adolescents had received reinforcement with BCG after their index case had been diagnosed, whereas 16 (19%) had not received this reinforcement (data was not documented in 12 (15%) of the study cohort). The frequency of anti-NDO-LID antibodies was higher (Table 4, p-value < 0.01), and elevated titers for those detected by Protein A (p-value =0.006), IgM (p-value = 0.016) and IgG (p-value = 0.007), were observed in individuals that had received reinforcement with BCG (Fig. 2b). Thus, a positive BCG immunization status was related to both lower rates and levels of M. leprae infection but lack of BCG presented as a risk.

It is well documented that contacts of multibacillary (MB) cases are at elevated risk of developing leprosy. We found that the majority (66; 80.5%) of the study subjects lived with a MB index case while 16 (19.5%) lived with a paucibacillary (PB) index case. Surprisingly, anti-NDO-LID antibodies detected by Protein A, IgM or IgG showed similar rates between MB or PB cases (p-value > 0.05). Similarly, differences were not observed in the overall antibody levels between children living with either MB or PB cases (p-value > 0.05, Fig. 3a). Therefore, in our evaluations the antibody responses of the children did not appear to be influenced by the clinical form of their respective index case.

Although the frequency of M. leprae infection was not impacted to the clinical form of the leprosy index case, we queried if the length of exposure to the leprosy index case could have an impact. Forty-five (54.9%) children and adolescents reported an exposure of less than 10 years while 37 (45.1%) reported exposure periods longer than 10 years (Table 4). When analyzing the presence of circulating antibodies, we found that anti-NDO-LID IgM levels were higher in individuals with > 10 years of exposure p-value = 0.044) (Fig. 3b). These data support the hypothesis that prolonged exposure to a leprosy index case increases the risk of M. leprae infection.

A binomial logistic analysis indicated that the single variable that most strongly associated with the frequency of M. leprae infection and increase antibody titers in children and adolescents in Colombia was residency in a geographic region linked to extreme poverty, such as the UAChR shared region (IgM/P- value = 0.004, Protein A/P-value = 0.045). Furthermore, it is probable that consumption of armadillo meat, and prolonged exposure to a leprosy case can be related with increase the risk to infection to M. leprae (Table 5).

The assessed socio-environmental variables also corresponded with antibody titers when assessed by a multiple lineal analysis, confirming that these variables behave like predictors and do indeed correspond with anti-NDO-LID levels (Protein A (R = 0.59, R² = 0.348, p-value = 0.0001), IgM (R = 0.746, R² = 0.528, p-value =0.0001) and IgG (R = 0.518, R² = 0.268, p-value = 0.0001).

The variables related to an increase in antibody titers in children and adolescents in the three endemic regions studied were: 1. residency in a geographic region linked to extreme poverty, such as the UAChR shared region (p-values < 0.0001 for IgM and 0.001 for Protein A); 2. > 10 years exposure to a leprosy case (p-values 0.018 for IgM and 0.027 for IgG); and 3. Consumption of armadillo meat (p-value 0.006 for IgM) (Table 6).

These data allowed us to create a series of equations based on the algebraic formula of the multiple lineal regression model, to predict anti-NDO-LID antibody titers (implying M. leprae infection) in the children and adolescent populations of the Colombia regions studied. The collinearity of the model was tested using VIF (variance inflation factor), coefficient of determination R², and residual analysis (Additional file 1). The model could be useful to predict IgM anti-NDO-LID antibody titers in the leprosy endemic regions evaluated. The model cannot be validated for Protein A and IgG anti NDO-LID.

Y: dependent variable, x: independent variable, β₀: coefficient (constant) of the dependent variable, β₁, β₂: coefficient that signifies a change in the dependent variable when the independent variable is present.