Artificial neural network to predict the effect of obesity on the risk of tuberculosis infection

  • Alaa Badawi
    Public Health Risk Sciences Division, Public Health Agency of Canada, Toronto; Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Canada.
    https://orcid.org/0000-0002-9115-0025
  • Christina J. Liu
    Department of Pharmacology and Toxicology, Faculty of Medicine, University of Toronto, Canada.
    https://orcid.org/0000-0001-9375-1361
  • Anas A. Rehim
    Ontario Tech University, Oshawa, Canada.
  • Alind Gupta
    Cytel Inc., Toronto, Canada.

ABSTRACT

Background: Body weight has been implicated as a risk factor for latent tuberculosis infection (LTBI) and the active disease.

Design and Methods:
 This study aimed to develop artificial neural network (ANN) models for predicting LTBI from body weight and other host-related disease risk factors. We used datasets from participants of the US-National Health and Nutrition Examination Survey (NHANES; 2012; n=5,156; 514 with LTBI and 4,642 controls) to develop three ANNs employing body mass index (BMI, Network I), BMI and HbA1C (as a proxy for diabetes; Network II) and BMI, HbA1C and education (as a proxy for socioeconomic status; Network III). The models were trained on n=1018 age- and sex-matched subjects equally distributed between the control and LTBI groups. The endpoint was the prediction of LTBI.

Results
: When data was adjusted for age, sex, diabetes and level of education, odds ratio (OR) and 95% confidence intervals (CI) for risk of LTBI with increased BMI was 0.85 (95%CI: 0.77 – 0.96, p=0.01). The three ANNs had a predictive accuracy varied from 75 to 80% with sensitivities ranged from 85% to 94% and specificities of approximately 70%. Areas under the receiver operating characteristic curve (AUC) were between 0.82 and 0.87. Optimal ANN performance was noted using BMI as a risk indicator.

Conclusion: 
Body weight can be employed in developing artificial intelligence-based tool to predict LTBI. This can be useful in precise decision making in clinical and public health practices aiming to curb the burden of tuberculosis, e.g., in the management and monitoring of the tuberculosis prevention programs and to evaluate the impact of healthy weight on tuberculosis risk and burden.

REFERENCES

GBD Tuberculosis Collaborators. Global, regional, and national burden of tuberculosis, 1990-2016: results from the Global Burden of Diseases, Injuries, and Risk Factors 2016 Study. Lancet Infect Dis 2018;18:1329-49. DOI: https://doi.org/10.1016/S1473-3099(18)30625-X

GBD Tuberculosis Collaborators. The global burden of tuberculosis: results from the Global Burden of Disease Study 2015. Lancet Infect Dis 2018;18:261–84. DOI: https://doi.org/10.1016/S1473-3099(17)30703-X

Kahwati LC, Feltner C, Halpern M, et al. Screening for latent tuberculosis infection in adults: An evidence review for the U.S. Preventive Services Task Force. Agency for Healthcare Research and Quality (US). Evidence Syntheses, No. 142; 2016. Accessed: 18 July 2020. Available from: https://www.ncbi.nlm.nih.gov/books/NBK385124/

Getahun H, Matteelli A, Chaisson RE, et al. Latent Mycobacterium tuberculosis infection. N Engl J Med 2015; 372:2127–35. DOI: https://doi.org/10.1056/NEJMra1405427

Uplekar M, Weil D, Lönnroth K, et al. WHO’s new End TB Strategy. Lancet 2015;385:1799-801. DOI: https://doi.org/10.1016/S0140-6736(15)60570-0

Dye C, Glaziou P, Floyd K, et al. Prospects for tuberculosis elimination. Annu Rev Public Health 2013;34:271-86. DOI: https://doi.org/10.1146/annurev-publhealth-031912-114431

Tverdal A. Body mass index and incidence of tuberculosis. Eur J Respir Dis 1986;69:355-62.

Cegielski JP, McMurray DN. The relationship between malnutrition and tuberculosis: Evidence from studies in humans and experimental animals. Int J Tuberc Lung Dis 2014;8:286-98.

Lönnroth K, Williams BG, Cegielski P, et al. A consistent log-linear relationship be¬tween tuberculosis incidence and body mass index. Int J Epidemiol 2010;39:149-55. DOI: https://doi.org/10.1093/ije/dyp308

Badawi A, Gregg B, Vasileva D. Systematic analysis for the relationship between obesity and tuberculosis. Pub Health 2020;186:246-56. DOI: https://doi.org/10.1016/j.puhe.2020.06.054

Roth J, Sahota N, Patel P, et al. Obesity paradox, obesity orthodox, and the metabolic syndrome: An approach to unity. Mol Med 2016; 2:873-85. DOI: https://doi.org/10.2119/molmed.2016.00211

Nuttall FQ. Body mass index: Obesity, BMI, and health: A critical review. Nutr Today 2015;50:117-28. DOI: https://doi.org/10.1097/NT.0000000000000092

Critchley JA, Restrepo BI, Ronacher K, et al. Defining a research agenda to address the converging epidemics of tuberculosis and diabetes: Part 1: Epidemiology and clinical management. Chest 2017;152:165-73. DOI: https://doi.org/10.1016/j.chest.2017.04.155

Lönnroth K, Roglic G, Harries AD. Improving tuberculosis prevention and care through addressing the global diabetes epidemic: From evidence to policy and practice. Lancet Diabetes Endocrinol 2014;2:730-9. DOI: https://doi.org/10.1016/S2213-8587(14)70109-3

Riza AL, Pearson F, Ugarte-Gil C, et al. Clinical management of concurrent diabetes and tuberculosis and the implications for patient services. Lancet Diabetes Endocrinol 2014;2:740-53. DOI: https://doi.org/10.1016/S2213-8587(14)70110-X

Badawi A, Sayegh S, Sallam M, et al. The global relationship between the prevalence of diabetes mellitus and incidence of tuberculosis: 2000-2012. Glob J Health Sci 2014;7:183-91. DOI: https://doi.org/10.5539/gjhs.v7n2p183

Cubilla-Batista I, Ruiz N, Sambrano D, et al. Overweight, obesity, and older age favor latent tuberculosis infection among household contacts in low tuberculosis-incidence settings within Panama. Am J Trop Med Hyg 2019;100:1141-4. DOI: https://doi.org/10.4269/ajtmh.18-0927

Zhang H, Li X, Xin H, et al. Association of body mass index with the tuberculosis infection: A population-based study among 17796 adults in rural China. Sci Rep 2017;7:41933. DOI: https://doi.org/10.1038/srep41933

WHO, European Respiratory Society. Digital health for the End TB Strategy: An agenda for action (WHO/HTM/TB/2015.21). 2015. Accessed: 19 June 2020. Available from: https://www.who.int/tb/areas-of-work/digital-health/Digital_health_EndTBstrategy.pdf

Doshi R, Falzon D, Thomas BV, et al. Tuberculosis control, and the where and why of artificial intelligence. ERJ Open Res 2017;3:56. DOI: https://doi.org/10.1183/23120541.00056-2017

Luger GF. Artificial intelligence: Structures and strategies for complex problem solving. 5th ed. Essex: Pearson Education Ltd.; 2005.

Cui S, Tseng HH, Pakela J, et al. Introduction to machine and deep learning for medical physicists. Med Phys 2020;47:e127‐47. DOI: https://doi.org/10.1002/mp.14140

Rajpurkar P, Irvin J, Ball RL, et al. Deep learning for chest radiograph diagnosis: A retrospective comparison of the CheXNeXt algorithm to practicing radiologists. PLoS Med 2018;15:e1002686. DOI: https://doi.org/10.1371/journal.pmed.1002686

Rajkomar A, Oren E, Chen K, et al. Scalable and accurate deep learning with electronic health records. NPJ Digit Med 2018;1:18. DOI: https://doi.org/10.1038/s41746-018-0029-1

Khan MT, Kaushik AC, Ji L, et al. Artificial neural networks for prediction of tuberculosis disease. Front Microbiol 2019;10:395. DOI: https://doi.org/10.3389/fmicb.2019.00395

Er O, Temurtas F, Tanrikulu AC. Tuberculosis disease diagnosis using artificial neural networks. J Med Syst 2010;34:299‐302. DOI: https://doi.org/10.1007/s10916-008-9241-x

Lakhani P, Sundaram B. Deep learning at chest radiography: Automated classification of pulmonary tuberculosis by using convolutional neural networks. Radiology 2017;284:574-82. DOI: https://doi.org/10.1148/radiol.2017162326

Mohadjer LMJ, Montaquila J, Waksberg J, et al. National Health and Nutrition Examination Survey III: Weighting and examination methodology. Hyattsville, MD; 1996. Available from: https://ceb.nlm.nih.gov/proj/dxpnet/nhanes/docs/doc/nhanes_analysis/wgt_exec.pdf

Barron MM, Shaw KM, Bullard KM, et al. Diabetes is associated with increased prevalence of latent tuberculosis infection: Findings from the National Health and Nutrition Examination Survey, 2011-2012. Diabetes Res Clin Pract 2018;139:366‐79. DOI: https://doi.org/10.1016/j.diabres.2018.03.022

Curtin LR, Mohadjer LK, Dohrmann SM, et al. National Health and Nutrition Examination Survey: sample design, 2007-2010. Vital Health Stat2 2013;160:1‐23.

Johnson CL, Dohrmann SM, Burt VL, Mohadjer LK. National health and nutrition examination survey: sample design, 2011-2014. Vital Health Stat2 2014;162:1‐33.

Mazurek GH, Jereb J, Vernon A, et al. Updated guidelines for using Interferon Gamma Release Assays to detect Mycobacterium tuberculosis infection – United States, 2010. MMWR Recomm Rep 2010;59:1‐25.

Miramontes R, Hill AN, Yelk Woodruff RS, et al. Tuberculosis infection in the United States: Prevalence estimates from the National Health and Nutrition Examination Survey, 2011-2012. PLoS One 2015;10:e0140881. DOI: https://doi.org/10.1371/journal.pone.0140881

American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diab Care 2010;33:S62-9. DOI: https://doi.org/10.2337/dc10-S062

Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and management of the metabolic syndrome: An American Heart Association/National Heart, Lung, and Blood Institute scientific statement. Circulation 2005;112:2735-52. DOI: https://doi.org/10.1161/CIRCULATIONAHA.105.169404

Brenner DR, Arora P, Garcia-Bailo B, et al. Plasma vitamin D levels and risk of metabolic syndrome in Canadians. Clin Invest Med 2011;34:E377. DOI: https://doi.org/10.25011/cim.v34i6.15899

Setayeshgar S, Whiting SJ, Vatanparast H. Prevalence of 10-year risk of cardiovascular diseases and associated risks in Canadian adults: The contribution of cardiometabolic risk assessment introduction. Int J Hyper 2013;2013:276564. DOI: https://doi.org/10.1155/2013/276564

Matthews DR, Hosker JP, Rudenski AS, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412‐9. DOI: https://doi.org/10.1007/BF00280883

Badawi A, Sayegh S, Sadoun E, et al. Relationship between insulin resistance and plasma vitamin D in adults. Diabetes Metab Syndr Obes 2014;7:297‐303. DOI: https://doi.org/10.2147/DMSO.S60569

CDC. National Health and Nutrition Examination Survey: 2011–2012 data documentation, codebook, and frequencies. Atlanta: National Center for Health Statistics; 2013.

Badawi A, Di Giuseppe G, Arora P. Cardiovascular disease risk in patients with hepatitis C infection: Results from two general population health surveys in Canada and the United States (2007-2017). PLoS One 2018;13:e0208839. DOI: https://doi.org/10.1371/journal.pone.0208839

Raschka S and Mirjalili V. Python Machine Learning, 2nd Edition. Birmingham: Packt Publishing Ltd.; 2017.

Chollet F. Xception: Deep learning with depthwise separable convolutions. arXiv 2017. 1610.02357v3. DOI: https://doi.org/10.1109/CVPR.2017.195

Valueva MV, Nagornov NN, Lyakhov PA, et al. Application of the residue number system to reduce hardware costs of the convolutional neural network implementation. Math Comp Simulation 2020;177:232-43. DOI: https://doi.org/10.1016/j.matcom.2020.04.031

Dupond S. A thorough review on the current advance of neural network structures. Ann Rev Control 2019;14:200-30.

Han SH, Kim KW, Kim S, et al. Artificial neural network: Understanding the basic concepts without mathematics. Dement Neurocogn Disord 2018;17:83-9. DOI: https://doi.org/10.12779/dnd.2018.17.3.83

He K, Zhang X, Ren S, Sun J. Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification. arXiv 2015;1502.01852. DOI: https://doi.org/10.1109/ICCV.2015.123

Ioffe S, Szegedy C. Batch normalization: Accelerating deep network training by reducing internal covariate shift. arXiv 2015;1502.03167.

Santurkar S, Tsipras D, Ilyas A, et al. How does batch normalization help optimization? arXiv 2019;1805.11604v5.

Nwankpa C, Ijomah W, Gachagan A, et al. Activation functions: Comparison of trends in practice and research for deep learning. arXiv 2018;1811.03378.

Goodfellow I, Bengio Y, Courville A. Deep Learning. Chapter 6: Deep feedforward Networks. MIT Press; 2016. pp. 164-223.

Riley RD, Ahmed I, Derbay TPA, et al. Summarising and validating test accuracy results across multiple studies for use in clinical practice. Stat Med 2015;34:2081-103. DOI: https://doi.org/10.1002/sim.6471

Trevethan R. Sensitivity, specificity, and predictive values: Foundations, pliabilities, and pitfalls in research and practice. Front Publ Health 2017;5:307. DOI: https://doi.org/10.3389/fpubh.2017.00307

Denisko D, Hoffman MM. Classification and interaction in random forests. Proc Natl Acad Sci USA 2018; 15:1690-2. DOI: https://doi.org/10.1073/pnas.1800256115

Cortes C, Vapnik V. Support-vector networks. Mach Learn 1995;20:273-97. DOI: https://doi.org/10.1007/BF00994018

Tolles J, Meurer WJ. Logistic regression: Relating patient characteristics to outcomes. JAMA 2016;316:533–4. DOI: https://doi.org/10.1001/jama.2016.7653

Yu B, Kumbier K. Artificial intelligence and statistics. Front Inf Technol Electronic Eng 2018;19:6-9. DOI: https://doi.org/10.1631/FITEE.1700813

Cirillo D, Catuara-Solarz S, Morey C, et al. Sex and gender differences and biases in artificial intelligence for biomedicine and healthcare. NPJ Digit Med 2020;3:81. DOI: https://doi.org/10.1038/s41746-020-0288-5

Zhang X, Jia H, Liu F, et al. Prevalence and risk factors for latent tuberculosis infection among health care workers in China: A cross-sectional study. PLoS One 2013;8:e66412. DOI: https://doi.org/10.1371/journal.pone.0066412

Sarivalasis A, Zellweger JP, Faouzi M, et al. Factors associated with latent tuberculosis among asylum seekers in Switzerland: a cross-sectional study in Vaud County. BMC Infect Dis 2012;12:285. DOI: https://doi.org/10.1186/1471-2334-12-285

Lule SA, Mawa PA, Nkurunungi G, et al. Factors associated with tuberculosis infection, and with anti-mycobacterial immune responses, among five-year old BCG-immunised at birth in Entebbe, Uganda. Vaccine 2015;33:796‐804. DOI: https://doi.org/10.1016/j.vaccine.2014.12.015

Martínez-Aguilar G, Serrano CJ, Castañeda-Delgado JE, et al. Associated risk factors for latent tuberculosis infection in subjects with diabetes. Arch Med Res 2015;46:221‐7. DOI: https://doi.org/10.1016/j.arcmed.2015.03.009

Kizza FN, List J, Nkwata AK, et al. Prevalence of latent tuberculosis infection and associated risk factors in an urban African setting. BMC Infect Dis 2015;15:165. DOI: https://doi.org/10.1186/s12879-015-0904-1

Leung CC, Lam TH, Chan WM, et al. Lower risk of tuberculosis in obesity. Arch Intern Med 2007;167:1297‐304. DOI: https://doi.org/10.1001/archinte.167.12.1297

Prospective Studies Collaboration, Whitlock G, Lewington S, et al. Body-mass index and cause-specific mortality in 900,000 adults: collaborative analyses of 57 prospective studies. Lancet 2009;373:1083‐96. DOI: https://doi.org/10.1016/S0140-6736(09)60318-4

Pednekar MS, Hakama M, Hebert JR, et al. Association of body mass index with all-cause and cause-specific mortality: Findings from a prospective cohort study in Mumbai (Bombay), India. Int J Epidemiol 2008;37:524‐35. DOI: https://doi.org/10.1093/ije/dyn001

Qin ZZ, Sander MS, Rai B et al. Using artificial intelligence to read chest radiographs for tuberculosis detection: A multi-site evaluation of the diagnostic accuracy of three deep learning systems. Sci Rep 2019;9:15000. DOI: https://doi.org/10.1038/s41598-019-51503-3

Da Costa LA, Arora P, García-Bailo B, et al. The association between obesity, cardiometabolic disease biomarkers, and innate immunity-related inflammation in Canadian adults. Diabetes Metab Syndr Obes 2012;5:347‐55. DOI: https://doi.org/10.2147/DMSO.S35115

Park HS, Park JY, Yu R. Relationship of obesity and visceral adiposity with serum concentrations of CRP, TNF-alpha and IL-6. Diabetes Res Clin Pract 2005;69:29‐35. DOI: https://doi.org/10.1016/j.diabres.2004.11.007

Karlsson EA, Beck MA. The burden of obesity on infectious disease. Exp Biol Med (Maywood) 2010;235:1412‐24. DOI: https://doi.org/10.1258/ebm.2010.010227

Lamas O, Marti A, Martínez JA. Obesity and immunocompetence. Eur J Clin Nutr 2002;56:S42‐5. DOI: https://doi.org/10.1038/sj.ejcn.1601484

Philips L, Visser J, Nel D, et al. The association between tuberculosis and the development of insulin resistance in adults with pulmonary tuberculosis in the Western sub-district of the Cape Metropole region, South Africa: A combined cross-sectional, cohort study. BMC Infect Dis 2017;17:570. DOI: https://doi.org/10.1186/s12879-017-2657-5

Thomson S. Achievement at school and socioeconomic background – an educational perspective. npj Science Learn 2018;3:5. DOI: https://doi.org/10.1038/s41539-018-0022-0

Olson NA, Davidow AL, Winston CA, et al. A national study of socioeconomic status and tuberculosis rates by country of birth, United States, 1996-2005. BMC Publ Health 2012;12:365. DOI: https://doi.org/10.1186/1471-2458-12-365

Cantwell MF, McKenna MT, McCray E, et al. Tuberculosis and race/ethnicity in the United States: Impact of socioeconomic status. Am J Respir Crit Care Med 1998;157:1016-20. DOI: https://doi.org/10.1164/ajrccm.157.4.9704036

Barr RG, Diez-Roux AV, Knirsch CA, et al. Neighborhood poverty and the resurgence of tuberculosis in New York City, 1984–1992. Am J Publ Health 2001;91:1487-93. DOI: https://doi.org/10.2105/AJPH.91.9.1487

Oliveira AL. Biotechnology, Big data and artificial intelligence. Biotechnol J 2019;14:1800613. DOI: https://doi.org/10.1002/biot.201800613

Ginsburg GS, Phillips KA. Precision medicine: from science to value. Health Aff 2018;37:694-701. DOI: https://doi.org/10.1377/hlthaff.2017.1624

Fan W, Bifet A. Mining big data: current status, and forecast to the future. ACM SIGKDD Explor Newsl 2014;16:1-5.

Caruana R, Lawrence S, Giles L. Overfitting in neural nets: backpropagation, conjugate gradient, and early stopping. In: Proc 13th Int Conf Neural Info Process Sys 2000; MIT Press, Cambridge, MA, USA. pp. 381-7. DOI: https://doi.org/10.1109/IJCNN.2000.857823

Badawi A, Drebot M, Ogden NH. Convergence of chronic and infectious diseases: a new direction in public health policy. Can J Publ Health 2019;110:523‐4. DOI: https://doi.org/10.17269/s41997-019-00228-x

Barajas A, Ochoa S, Obiols JE et al. Gender differences in individuals at high-risk of psychosis: A comprehensive literature review. Sci World J 2015;2015:430735. DOI: https://doi.org/10.1155/2015/430735

Obermeyer Z, Powers B, Vogeli C, et al. Dissecting racial bias in an algorithm used to manage the health of populations. Science 2019;366:447-53. DOI: https://doi.org/10.1126/science.aax2342

James WP, Ferro-Luzzi A, Waterlow JC. Definition of chronic energy deficiency in adults. Report of a working party of the International Dietary Energy Consultative Group. Eur J Clin Nutr 1988;42:969‐81.

Moreira-Teixeira L, Mayer-Barber K, Sher A, et al. Type I interferons in tuberculosis: Foe and occasionally friend. J Exp Med 2018;215:1273‐85. DOI: https://doi.org/10.1084/jem.20180325

Procaccini C, Lourenco EV, Matarese G, et al. Leptin signaling: A key pathway in immune responses. Curr Signal Transduct Ther 2009;4:22‐30. DOI: https://doi.org/10.2174/157436209787048711

Odone A, Houben RM, White RG, et al. The effect of diabetes and undernutrition trends on reaching 2035 global tuberculosis targets. Lancet Diabet Endocrinol 2014;2:754‐64. DOI: https://doi.org/10.1016/S2213-8587(14)70164-0