SETTING: Dar es Salaam, Tanzania. OBJECTIVE: To determine the prevalence of latent tuberculosis (TB) infection (LTBI) among adolescents in a country with a high TB burden, and examine risks of LTBI according to their social activity patterns. METHODS: A cross-sectional study nested within a phase 2b randomised, placebo controlled, double blind study and consisted of 824 adolescents, 13 - 15 years old who had received Bacillus Calmette-Guérin (BCG) vaccine, were attending public secondary schools and had no evidence of active tuberculosis (TB). Anthropometric measurements were obtained, a questionnaire administered, and phlebotomy performed for a T spot interferon- γ release assay (IGRA) to detect LTBI. RESULTS: Among 824 subjects, 149 (18%) had a positive IGRA. After adjusting for the influence of household socioeconomic status, history of TB contact, living environment and nutritional status, LTBI risk was higher in subjects with than without regular informal encounters with traditional alcoholic beverage drinkers (AOR, 6.37 [1.84 - 22.00]). Other significant factors for LTBI risk included contact with TB patient at school (AOR, 3.34 [1.14 - 9.80]), and living close to a health facility, as was observed among those from houses within a 10 - 30-minute walking distance to the nearest health facility, who were less likely to be IGRA-positive than those who were living within a 10-minute walking distance (AOR, 0.30 [95%CI, 0.13 - 0.69]). CONCLUSION: This IGRA study revealed a high prevalence of LTBI among adolescents in Dar es Salaam, Tanzania with prior BCG immunization. Informal social encounters were identified as independent risk factors for LTBI, along with a history of contact with TB patients, living environment characteristics and household socioeconomic status. Efforts focusing on risk of MTB transmission in adolescents at informal social gatherings will improve interventions to reduce LTBI in this population and consequently the subsequent risk of developing active TB disease.
Despite a decline in its global prevalence, tuberculosis (TB) still causes more deaths than other infectious diseases [
About one third of the world’s population is infected with the causative agent of TB, Mycobacterium tuberculosis (MTB) [
Data on the prevalence of LTBI and risk factors for LTBI in Tanzania are more limited, and are not available for adolescents. This is true despite the high incidence of TB disease among children <14 years old in Tanzania [
The T spot interferon-γ release assay (IGRA) is an advanced TB diagnostic tool, and is more specific than the tuberculin skin testing (TST) in identifying LTBI among BCG-vaccinated individual [
In this study, we used IGRA as the most specific approach to detect LTBI in adolescents at school, who had received BCG at birth. Our objectives were to obtain an accurate estimate of the prevalence of LTBI in adolescents, and to identify risk factors for LTBI with a focus on both the home environment and social encounters outside the home.
This study was conducted between April 6 and September 23, 2016, in Dar es Salaam, Tanzania, with a population of 5,465,420 as of 2016 [
Subjects were recruited from 16 secondary schools chosen from a list of 139 public non-boarding secondary schools within the Dar es Salaam region [
Initial screening for the study was performed after obtaining written informed consent and assent from a total of 936 subjects. The inclusion criteria were: age between 13 and 15 years, attendance at a public non-boarding secondary school, evidence of BCG vaccination (either the presence of a scar or documentation) and the absence of signs of active TB or symptoms suggesting active TB disease based on the Tanzanian TB screening questionnaire [
Body mass index (BMI) (kg/m2) was calculated based on height and weight measurements. BMI-for-age percentiles for both genders, obtained using the Centres for Disease Control (CDC) growth charts [
The IGRA was performed using the T-SPOT.TB (Oxford Immunotec, Oxford, UK) assay according to the manufacturer’s instructions. Venous blood samples were analysed at the Tuberculosis Research Institute at Muhimbili University of Health and Allied Sciences (TRIM-TB) in Dar es Salaam, Tanzania [
A questionnaire was developed at the Department of Global Health Entrepreneurship of Tokyo Medical and Dental University, and included questions about prior active TB, contact with TB patients, social activity patterns, living environment and household socioeconomic characteristics. The original English version questionnaire was translated into Swahili, which was then back translated to English for confirmation.
After receiving instructions on how to complete the questionnaire, the adolescents were allowed to take the questionnaire home to answer all questions with the help of their parents/guardians. At the time of submission, nurses inspected the completed questionnaires to confirm that they had been filled out correctly. To ensure consistency in instruction and interviews, all nurses involved in the study received a full day of training before initiation of the study.
We obtained the subject’s past history of diagnosis and treatment for active TB disease, household contact with a TB patient, knowledge of a neighbour that had TB, and knowledge of a fellow student diagnosed with TB.
To understand the social activity patterns of the adolescents, the questionnaire inquired about the means of transport used to attend school, encounters with traditional alcoholic beverage drinkers, visits to market places and walking distance between the house and the nearest health facility. All these activities increase risks for contact with active TB patients.
In relation to the living environment characteristics, the number of living rooms and bedrooms, brightness from sunlight in living rooms and bedrooms and number of older family members living in the same house were determined.
The education levels of the father and mother, employment of father and household wealth status evaluated according to the Demographic and Health Survey (DHS) guidelines [
Statistical analyses were performed using SPSS Statistics v22.0 (IBM Corp., Armonk, NY, USA). Percentage distributions according to IGRA results were calculated for 824 subjects. After excluding 20 subjects with indeterminate and borderline results, the percentages of IGRA-positive cases according to their characteristics were evaluated using cross-tabulation with the χ2 test. After excluding 75 subjects with prior history of TB diagnosis (23 who responded “Yes” and 52 who responded “I do not remember”), logistic regression analysis to examine the association between IGRA positivity and characteristics of the subjects were performed with three models: 1) crude; 2) adjusted by wealth, history of contact with TB patients in their households and in the community; and 3) by wealth, history of contact with TB patients in their households, in the community, living conditions and nutritional status. All missing data were excluded from multiple logistic regression analyses models. In all analyses, P < 0.05 indicated statistical significance.
This study was approved by the Institution Review Boards of the Medical Schools of Tokyo Medical and Dental University, Japan; Muhimbili University of Health and Allied Sciences, Tanzania; and Geisel School of Medicine at Dartmouth, NH, USA. Permission to recruit secondary school students was obtained from the Ministry of Health, Community Development, Gender, Elderly and Children, the Ministry of Education and Vocational Training, and the President’s Office for Regional Administration and Local Government of Tanzania.
CATEGORY | n | % |
---|---|---|
IGRA Result | ||
Positive | 149 | 18.1 |
Negative | 655 | 79.5 |
Indeterminate | 19 | 2.3 |
Borderline | 1 | 0.1 |
Total | 824 | 100.0 |
n, number of subjects; %, percentage.
N = 804 | IGRA-positive adolescents | |||||
---|---|---|---|---|---|---|
% | n | % | P | |||
Gender and nutritional status | ||||||
Gender | 0.22 | |||||
Boys | 358 | 44.5 | 73 | 20.4 | ||
Girls | 446 | 55.4 | 76 | 17.0 | ||
Body Mass Index (BMI) percentile | 0.12 | |||||
<5% [underweight] | 23 | 2.9 | 6 | 26.1 | ||
5% - 85% [healthy] | 591 | 73.5 | 110 | 18.6 | ||
85% - 95% [overweight] | 124 | 15.4 | 16 | 12.9 | ||
>95% [obese] | 66 | 8.21 | 17 | 25.8 | ||
History of TB and contact with TB patients | ||||||
History of being treated for TB | 0.01 | |||||
No | 729 | 90.7 | 132 | 18.1 | ||
Yes | 23 | 2.9 | 9 | 39.1 | ||
Unknown | 52 | 6.5 | ||||
Contact with TB patient(s) in household | 0.03 | |||||
No | 616 | 76.6 | 100 | 16.2 | ||
Yes | 90 | 11.2 | 23 | 25.6 | ||
Unknown | 98 | 12.2 | ||||
Contact with TB patient(s) in community | <0.01 | |||||
No | 493 | 61.3 | 83 | 16.8 | ||
Yes | 59 | 7.3 | 25 | 42.4 | ||
Unknown | 252 | 31.3 | ||||
Contact with TB patient(s) at school | <0.01 | |||||
No | 663 | 82.5 | 111 | 16.7 | ||
Yes | 34 | 4.2 | 14 | 41.2 | ||
Unknown | 107 | 13.3 | ||||
Social activity patterns among adolescents | ||||||
Means of transport to and from school | 0.04 | |||||
Public transport | 308 | 38.3 | 46 | 14.9 | ||
Walking | 476 | 59.2 | 99 | 20.8 | ||
Bicycle | 20 | 2.5 | 5 | 25.0 |
Encounter with traditional alcoholic beverage drinkers | 0.01 | |||||
---|---|---|---|---|---|---|
No, never | 563 | 70.0 | 101 | 17.9 | ||
Yes, only few times in the past | 142 | 17.7 | 32 | 22.5 | ||
Yes, once to a few times monthly | 41 | 5.1 | 2 | 4.9 | ||
Yes, once a week or more frequently | 32 | 4.0 | 11 | 34.4 | ||
Unknown | 26 | 3.2 | ||||
Visits to market places | 0.01 | |||||
Once or less per month | 411 | 51.1 | 63 | 15.3 | ||
More than once per month | 360 | 44.8 | 82 | 22.8 | ||
Unknown | 33 | 4.1 | ||||
Closeness to health facility | 0.04 | |||||
<10 minutes | 286 | 35.6 | 68 | 23.8 | ||
10 - 30 minutes | 197 | 24.5 | 27 | 13.7 | ||
>30 minutes | 96 | 11.9 | 17 | 17.7 | ||
Unknown | 225 | 28.0 | ||||
Living environments | ||||||
Number of living room(s) | <0.01 | |||||
0 | 10 | 1.2 | 1 | 10.0 | ||
1 | 268 | 33.3 | 70 | 26.1 | ||
2 | 283 | 35.2 | 39 | 13.8 | ||
Unknown | 243 | 30.2 | ||||
Brightness from sunlight in living room(s) | <0.01 | |||||
Lack of sunlight | 13 | 1.6 | 2 | 15.4 | ||
Somewhat bright | 438 | 54.5 | 100 | 22.8 | ||
Very bright | 222 | 27.6 | 22 | 9.9 | ||
Unknown | 131 | 16.3 | ||||
Number of bedrooms(s) | <0.01 | |||||
1 | 122 | 15.2 | 16 | 13.1 | ||
2 | 134 | 16.7 | 38 | 28.4 | ||
3 | 354 | 44.0 | 44 | 12.4 | ||
≥4 | 94 | 11.7 | 32 | 34.0 | ||
Unknown | 100 | 12.4 |
Brightness from sunlight in bedroom(s) | 0.01 | |||||
---|---|---|---|---|---|---|
Lack of sunlight at all | 21 | 2.6 | 3 | 14.3 | ||
Somewhat bright | 540 | 67.2 | 113 | 20.9 | ||
Very bright | 154 | 19.2 | 16 | 10.4 | ||
Unknown | 89 | 11.0 | ||||
Family members older than adolescent | 0.04 | |||||
1 | 61 | 7.6 | 4 | 6.6 | ||
2 | 138 | 17.2 | 31 | 22.5 | ||
≥3 | 535 | 66.5 | 99 | 18.5 | ||
Unknown | 70 | 8.7 | ||||
Household socioeconomic characteristics | ||||||
Father’s educational attainment | <0.01 | |||||
No | 21 | 2.6 | 12 | 57.1 | ||
Primary education | 288 | 35.8 | 69 | 24.0 | ||
Higher education | 305 | 37.9 | 43 | 14.1 | ||
Unknown | 190 | 23.6 | ||||
Mother’s educational attainment | 0.03 | |||||
No | 17 | 2.1 | 3 | 17.6 | ||
Primary education | 370 | 46.0 | 89 | 24.1 | ||
Higher education | 238 | 29.6 | 35 | 14.7 | ||
Unknown | 179 | 22.3 | ||||
Father’s employment | 0.01 | |||||
Unemployed | 29 | 3.6 | 1 | 3.4 | ||
Businessman | 448 | 55.7 | 94 | 21.0 | ||
Works in the office | 95 | 11.8 | 8 | 8.4 | ||
Industrial worker | 39 | 4.9 | 10 | 25.6 | ||
Unknown | 193 | 24.0 | ||||
Wealth status of household | 0.01 | |||||
Richest | 162 | 20.1 | 17 | 10.5 | ||
Rich | 161 | 20.0 | 35 | 21.7 | ||
Middle | 158 | 19.7 | 27 | 17.1 | ||
Poor | 165 | 20.5 | 43 | 26.1 | ||
Poorest | 158 | 19.7 | 27 | 17.1 |
TB, Tuberculosis; N, number of subjects after excluding those with borderline and indeterminate IGRA results; %, percentage; n, number of subjects with positive IGRA results; P, P-value; ≥, more or equal to; ≤, less or equal to; <, less than; >, more than.
Characteristics | n | MODEL 1a | P | MODEL 2b | P | MODEL 3c | P | |
---|---|---|---|---|---|---|---|---|
COR [95%CI] | AOR [95%CI] | AOR [95%CI] | ||||||
History of contact with TB patients | ||||||||
Contact with TB patient(s) in household | ||||||||
No | 569 | Ref | Ref | Ref | ||||
Yes | 74 | 1.24 [0.66 - 2.32] | 0.49 | 1.16 [0.60 - 2.27] | 0.66 | 1.26 [0.54 - 2.92] | 0.60 | |
Contact with TB patient(s) in the community | ||||||||
No | 463 | Ref | Ref | Ref | ||||
Yes | 51 | 2.41 [1.26 - 4.62] | 0.01 | 1.92 [0.93 - 3.97] | 0.08 | 2.02 [0.84 - 4.88] | 0.12 | |
Contact with TB patient(s) at school | ||||||||
No | 603 | Ref | Ref | Ref | ||||
Yes | 35 | 3.57 [1.71 - 7.48] | <0.01 | 3.55 [1.54 - 8.17] | <0.01 | 3.34 [1.14 - 9.80] | 0.03 | |
Social activity patterns of adolescents | ||||||||
Encounter with traditional alcoholic beverage drinkers | ||||||||
Never | 522 | Ref | Ref | Ref | ||||
Yes, a few times in the past | 119 | 0.85 [0.50 - 1.47] | 0.57 | 0.72 [0.37 - 1.40] | 0.33 | 0.26 [0.09 - 0.81] | 0.02 | |
Yes, once to a few times a month | 38 | 0.24 [0.06 - 1.02] | 0.05 | 0.32 [0.07 - 1.38] | 0.13 | 0.40 [0.09 - 1.79] | 0.23 | |
Yes, once a week to more frequently | 29 | 3.00 [1.35 - 6.66] | 0.01 | 2.90 [1.15 - 7.34] | 0.03 | 6.37 [1.84 - 22.00] | <0.01 | |
Closeness to health facility | ||||||||
≤10 mins | 259 | Ref | Ref | Ref | ||||
10 - 30 mins | 177 | 0.57 [0.34 - 0.96] | 0.03 | 0.44 [0.25 - 0.78] | 0.01 | 0.30 [0.13 - 0.69] | 0.01 | |
≥30 minutes | 95 | 0.81 [0.44 - 1.49] | 0.50 | 0.50 [0.23 - 1.06] | 0.07 | 0.58 [0.23 - 1.46] | 0.25 | |
Living environment | ||||||||
Number of living rooms | ||||||||
1 | 238 | Ref | Ref | Ref | ||||
2 | 263 | 0.42 [0.26 - 0.67] | <0.01 | 0.46 [0.26 - 0.80] | 0.01 | 0.55 [0.30 - 1.00] | 0.05 | |
Brightness from sunlight in living room(s) | ||||||||
Very bright | 206 | Ref | Ref | Ref | ||||
Somewhat bright | 398 | 2.72 [1.60 - 4.63] | <0.01 | 2.17 [1.22 - 3.86] | 0.01 | 1.30 [0.61 - 2.80] | 0.50 | |
Number of bedrooms | ||||||||
1 | 107 | Ref | Ref | Ref | ||||
2 | 127 | 2.29 [1.15 - 4.56] | 0.02 | 1.91 [0.90 - 4.05] | 0.09 | 1.51 [0.49 - 4.64] | 0.48 | |
3 | 319 | 0.84 [0.43 - 1.62] | 0.60 | 0.68 [0.33 - 1.41] | 0.30 | 0.35 [0.11 - 1.06] | 0.06 | |
≥4 | 95 | 3.40 [1.67 - 6.91] | <0.01 | 3.88 [1.74 - 8.67] | <0.01 | 3.85 [1.12 - 13.20] | 0.03 |
Brightness from sunlight in bedrooms | ||||||||
---|---|---|---|---|---|---|---|---|
Very bright | 135 | Ref | Ref | Ref | ||||
Somewhat bright | 492 | 3.33 [1.69 - 6.58] | <0.01 | 2.89 [1.39 - 6.00] | <0.01 | 2.02 [0.74 - 5.50] | 0.17 | |
Household socioeconomic characteristics | ||||||||
Father’s educational attainment | ||||||||
No education | 18 | Ref | Ref | Ref | ||||
Primary Education | 264 | 0.43 [0.16 - 1.21] | 0.11 | 3.10 [0.34 - 28.07] | 0.32 | 2.43 [0.24 - 24.13] | 0.45 | |
Higher education | 276 | 0.18 [0.06 - 0.52] | <0.01 | 1.66 [0.18 - 15.10] | 0.65 | 1.20 [0.12 - 11.89] | 0.87 | |
Mother’s educational attainment | ||||||||
Ref: No education | 15 | Ref | Ref | Ref | ||||
Primary education | 342 | 1.25 [0.34 - 4.54] | 0.74 | 1.54 [0.38 - 6.26] | 0.55 | 2.20 [0.23 - 20.71] | 0.49 | |
Higher education | 215 | 0.70 [0.19 - 2.63] | 0.60 | 1.13 [0.27 - 4.72] | 0.87 | 1.63 [0.17 - 15.77] | 0.67 | |
Wealth status of household | ||||||||
Richest quintile | 146 | Ref | Ref | Ref | ||||
Rich quintile | 140 | 1.92 [0.92 - 3.97] | 0.08 | 1.66 [0.76 - 3.63] | 0.20 | 1.69 [0.63 - 4.53] | 0.29 | |
Middle quintile | 142 | 2.25 [1.10 - 4.61] | 0.03 | 1.70 [0.77 - 3.72] | 0.19 | 2.55 [1.00 - 6.48] | 0.05 | |
Poor quintile | 157 | 3.91 [2.00 - 7.64] | <0.01 | 3.35 [1.67 - 6.72] | <0.01 | 2.51 [1.02 - 6.19] | 0.05 | |
Poorest quintile | 144 | 2.35 [1.16 - 4.76] | 0.02 | 1.71 [0.79 - 3.71] | 0.18 | 1.05 [0.35 - 3.18] | 0.93 |
COR, crude odds ratio; AOR, adjusted odds ratio; P, P-value; Ref, reference; 95%CI, 95% Confidence Interval; ≥, more or equal to; ≤, less or equal to; <, less than; >, more than. aUnadjusted model; bModel adjusted for wealth index, household and community tuberculosis contact; cModel adjusted for wealth index, household and community tuberculosis contact, number of living rooms, brightness in the living room and bedroom as well as BMI percentiles.
within a 10-minute walking distance (AOR, 0.30 [95%CI, 0.13 - 0.69]). Significant associations were also observed between IGRA positivity and living in a house with one living room, living in a house with four or more bedrooms, contact with a TB patient at school and wealth status.
This is the first study to use the IGRA to examine the prevalence of LTBI among in-school adolescents in Tanzania. Encounters with traditional alcoholic beverage drinkers and living environments were identified as risk factors for LTBI, in addition to household socioeconomic status and history of contact with TB patients in the household and community.
The LTBI prevalence rate of 18.1% identified by IGRA among Tanzanian adolescents 13 - 15 years old is lower than the prevalence rate of 41.7% reported for those 11 - 15 years old in Cape Town, South Africa [
Because most Tanzanian adolescents have received BCG [
LTBI was associated with a history of household contact with a TB patient, but this relationship disappeared after excluding subjects with history of prior TB. These findings suggested most subjects who had lived in the same household as a TB patient, might have later been diagnosed with active TB themselves, as postulated by the WHO [
Regarding exposure in the community, the risk of acquiring LTBI was higher among adolescents with than without a history of encounters with traditional alcoholic beverage drinkers. This observation indicated a role of informal gatherings within communities in the spread of MTB. Traditional alcoholic beverage drinking is usually done at gatherings that are associated with ideal environments for MTB transmission like small, overcrowded and poor ventilated traditional bars and contact with drinkers who are active TB patients. These drinkers in turn spread MTB to adolescents with whom they interact informally within the community [
Living in a house with four or more bedrooms was shown to increase the risk of LTBI, in contrast to previous studies [
This study also revealed an association between residing in a house located closest to a health facility and IGRA positivity. Residing close to a health facility might increase adolescents’ risks for contact with active TB patients who attend those facilities for TB treatment. Consistent with findings from previous studies [
Living in a house with two living rooms, compared to one living room, was associated with a lower risk of LTBI. This observation point to an increase in LTBI risk associated with poor ventilation inside the living rooms, supporting reports that overcrowding and poor ventilation increase the risk of acquiring MTB [
Our study had notable strengths including being the first study to use IGRA, rather than TST, to identify LTBI among BCG-vaccinated adolescents in Tanzania. Using IGRA also provided logistic advantages as it does not require a subsequent visit 72 hours from inoculation to interpret the test results, hence convenient and contributes to a lower loss-to-follow-up rate than TST. The TB risk factor questionnaire used ensured ability to identify local LTBI risk factors based on locally available evidence. Also, overall recruitment challenges and participation bias were minimized by avoiding mandatory HIV testing among participants as the test, if were mandatory, would have limited voluntary participation to avoid stigma and discrimination since HIV/AIDS is prevalent in Tanzania [
However, our study also had some limitations including omission of a small number of subjects with missing data while conducting the logistic regression analysis, but since their numbers were small, omitting them did not affect our results. In addition, being cross-sectional in nature, it failed to establish causality assumption. Moreover, our study excluded out-of-school adolescents, which limited generalizability of the findings to all adolescents in Tanzania.
This is the first study to use the IGRA to examine LTBI prevalence among in-school adolescents in Tanzania. Informal social encounters were identified as independent risk factors for LTBI, along with a history of contact with TB patients, living environment characteristics and household socioeconomic status. Efforts focusing on risk of MTB transmission in adolescents at informal social gatherings will improve interventions to reduce LTBI in this population and consequently the subsequent risk of developing active TB disease.
The authors gratefully acknowledge the assistance from Dr. Masashi Kizuki of Tokyo Medical and Dental University in designing the questionnaire. They also thank Dar-PIA team members at Dar-PIA Clinic in Dar es Salaam, Tanzania (Dr. Albert Katana and Dr. Maryam Amour for coordinating meetings at the Ministry of Education and Vocational Training (MoEVT), Ministry of Health, Community Development, Gender, Elderly and Children (MoHCDEC) and President’s Office-Regional Administration and Local Government (PO-RALG) as well as organising meetings with schools’ administration, Dr. Suleiman Chum, Sr. Mary Ngatoluwa, Sr. Asha Swaleh and Mr. Chijano Makunenge for collecting data and Ms. Betty Mchaki for conducting IGRA tests.
The Global Health Innovative Technology Fund (G2015-147) and Japan Society for Promotion of Science (17H02164) supported this work. The funders had no role in study design data collection and analysis, decision to publish, or preparation of the manuscript.
Maro, I.I., Nakamura, K., Seino, K., Pallangyo, K., Munseri, P., Matee, M. and von Reyn, C.F. (2018) Social Activity Patterns Drive High Rates of Latent Tuberculosis Infection among Adolescents in Urban Tanzania. Journal of Tuberculosis Research, 6, 81-95. https://doi.org/10.4236/jtr.2018.61008