Prediction of hormonal and metabolic disorders in young women with overweight and obesity: the effectiveness of artificial neural networks
Keywords:obesity, metabolic disorders, mathematical predictive model, discriminant analysis, logistic regression, artificial neural networks
Background. The relevance of the study is due to the need to create methodological approaches to the formation of risk groups for the development of metabolic syndrome — the basis of chronic noninfectious pathology of the cardiovascular system in young women with excess body weight of varying degrees. The purpose was to create models for predicting the development of metabolic syndrome in young women with overweight and obesity — representatives of the Ukrainian population using classical methods of statistical analysis (discriminant analysis, logistic regression) and artificial neural networks and comparative analysis of the prognostic accuracy of the created models. Materials and methods. One hundred and thirty women with average age of 28.64 ± 6.91 years were examined. They had exogenous constitutional excess body weight of varying degrees. Body mass index and waist circumference were determined. The method of bioimpedance analysis was used to evaluate body composition (fat, relative fat, fatfree, active cell body mass); enzyme immunoassay — to determine the levels of insulin and leptinemia. The HOMAIR was calculated. The secretion of melatonin was assessed by the level of its metabolite 6sulfatoxymelatonin in urine by Druex method modified by G.V. Zubkov; serotonin — by the fluorimetric method of V.I. Kulinsky and L.V. Kostyukovska. The presence of sleep and eating disorders was assessed by questionnaires. Results. According to the results of a comprehensive laboratory and instrumental examination using methods of discriminant analysis, logistic regression and artificial neural networks, four mathematical models were created that allow us to assess the risk of metabolic syndrome in young women with excess body weight of varying degrees — representatives of the Ukrainian population by the levels of anthropometric indicators, parameters of body composition, carbohydrate and lipid metabolism, indicators that characterize the features of melatonin secretion. Conclusions. Diagnostic characteristics of all proposed models (sensitivity, specificity, accuracy, odds ratio) are quite high, but the greatest diagnostic informativeness is determined for the model using artificial neural networks built in the program Statistica StatSoft. The use of such a model as a tool for determining the risk of developing the metabolic syndrome in specific young women with excess body weight of varying degrees in the practical healthcare system will improve the riskstratification of the metabolic syndrome and provide timely therapy to prevent its complications.
Misjura KV. Nadlyshkova massa tila i ozhyrinnja: populjacijni, klinichni ta gormonal'ni osoblyvosti, mehanizmy formuvannja metabolichnyh uskladnen'. Diss. dokt. med. nauk [Overweight and obesity: population, clinical and hormonal features, mechanisms of formation of metabolic complications. Dr. med. sci. diss.]. Kharkiv; 2018. 47 p. (in Ukrainian).
Nyberg ST, Batty GD, Pentti J, et al. Obesity and loss of disease-free years owing to major non-communicable diseases: a multicohort study. Lancet Public Health. 2018 Oct;3(10):e490-e497. doi: 10.1016/S2468-2667(18)30139-7.
World Health Organization (WHO). Obesity and overweight: Key facts. Available from: https://www.who.int/en/news-room/fact-sheets/detail/obesity-and-overweight. Accessed: February 16, 2018.
Lobykina EN. Organizatsiia profilaktiki i lecheniia ozhireniia i izbytochnoi massy tela vzroslogo naseleniia krupnogo promyshlennogo tsentra na primere g. Novokuznetska. Diss. dokt. med. nauk [Organization of the prevention and treatment of obesity and overweight of the adult population of a large industrial center on the example of Novokuznetsk. Dr. med. sci. diss.]. Kemerovo; 2009. 331 p. (in Russian).
Yang Y, Tian CH, Cao J, Huang XJ. Research on the application of health management model based on the perspective of mobile health. Medicine (Baltimore). 2019 Aug;98(33):e16847. doi: 10.1097/MD.0000000000016847.
Mitchenko OI, Romanov VJu, Javors'ka KO. High cardiovascular risk in patients with hypertension and obesity. Zdorov’ja Ukrai'ny. Kardiologija, Revmatologija, Kardiohirurgija. 2012;(23-24):24-25. (in Ukrainian).
Kovalenko VM, Talayeva TV, Kozliuk AS. Metabolic syndrome: mechanisms, value as a predictor of cardiovascular diseases, approaches in diagnosis and treatment. Ukrainian Journal of Cardiology.2013;(5):80-87. (in Ukrainian).
Shapo L, Pomerleau J, McKee M, Coker R, Ylli A. Body weight patterns in a country intransition: apopulation-based survey in Tirana City, Albania. Public Health Nutr. 2003 Aug;6(5):471-477. doi: 10.1079/PHN2002451.
Starodubov VI, Mikhailova IuV, Ivanova AE. Zdorov'e naseleniia Rossii v sotsial'nom kontekste 90-kh godov: problemy i perspektivy [The health of the population of Russia in the social context of the 90s: problems and prospects]. Moscow: Meditsina; 2003. 96 p. (in Russian).
Tishchuk EA, Shchepin VO. Zdravookhranenie v Rossii. XX vek [Health care in Russia. XX century]. Moscow; 2001. 48 p. (in Russian).
Svishhenko JeP, Bagrij AE, Jena LM, et al. Rekomendacii' Ukrai'ns'koi' Asociacii' kardiologiv z profilaktyky ta likuvannja arterial'noi' gipertenzii': posibnyk do Nacional'noi' programy profilaktyky i likuvannja arterial'noi' gipertenzii'. [Guidelines of the Ukrainian Association of Cardiologists for the prevention and treatment of hypertension: a guide to the National Program for the Prevention and Treatment of Hypertension]. Kyiv; 2008. 80 p. (in Ukrainian).
Kir'ianov BF, Tokmachev MC. Matematicheskie modeli v zdravookhranenii: monografiia [Mathematical models in healthcare: a monograph]. Velikij Novgorod; 2009. 277 p. (in Russian).
Borovikov VP. Statistica: iskusstvo analiza dannykh na komp'iutere [Statistica: the art of computer data analysis]. SPb: Piter; 2001. 700 p. (in Russian).
Dubrov AM, Mkhitarian VS, Troshin LI. Mnogomernye statisticheskie metody [Multivariate statistical methods]. Moscow: Finansy i statistika; 2000. 352 p. (in Russian).
Institute for Innovative Education and Youth Policy; Scientific-educational center of applied informatics of NAS of Ukraine. Naukovi doslidzhennja: perspektyvy innovacijnogo rozvytku suspil'stva i tehnologij: Materialy Mizhnarodnoi' naukovo-praktychnoi' konferencii. Odesa, 28-29 zhovtnja 2016 r. [Scientific research: prospects for innovative development of society and technology: Materials of the International scientific-practical conference. 2016, October 28-29; Odessa, Ukraine]. Odessa; 2016. 196 p. (in Ukrainian).
Andreichikov AV, Andreichikova ON. Intellektual'nye informatsionnye sistemy: uchebnik dlia vuzov [Intelligent information systems: university textbook]. Moscow: Finansy i statistika; 2004. 423 p. (in Russian).
Volchek YA, Shyshko VM, Spiridonova OS, Mokhort TV. Position of the model of the artificial neural network in medical expert systems. Juvenis Scientia. 2017;(9):4-9. doi: 10.15643/jscientia.2017.9.001. (in Russian).
Artificial neural network. Available from: https://en.wikipedia.org/wiki/Artificial_neural_network. Accessed: January 2, 2020.
Chornopys'ka Ju. Prediction using neural networks. Available from: https://wiki.tntu.edu.ua/Прогнозування_за_допомогою_нейронних_мереж. Accessed: March 20, 2012. (in Ukrainian).
STATISTICA Automated Neural Networks. Available from:
http://statsoft.ru/products/STATISTICA_Neural_Networks/. Accessed: September 23, 2019. (in Russian).
Callan D, Mills L, Nott C, England R, England S. A Tool for Classifying Individuals with Chronic Back Pain: Using Multivariate Pattern Analysis with Functional Magnetic Resonance Imaging Data. PLoS One. 2014 Jun 6;9(6):e98007. doi: 10.1371/journal.pone.0098007.
Toney L, Vesselle H. Neural networks for nodal staging of non-small cell lung cancer with FDG PET and CT: importance of combining uptake values and sizes of nodes and primary tumor. Radiology. 2014 Jan;270(1):91-8. doi: 10.1148/radiol.13122427.
Andersson B, Andersson R, Ohlsson M, Nilsson J. Prediction of severe acute pancreatitis at admission to hospital using artificial neural networks. Pancreatology. 2011;11(3):328-35. doi: 10.1159/000327903.
Gorunescu F, Gorunescu M, Saftoiu A, Vilmann P, Belciug S. Competitive/collaborative neural computing system for medical diagnosis in pancreatic cancer detection. Expert System. 2011;28(1):33-48. doi:10.1111/j.1468-0394.2010.00540.x.
Hirose H, Takayama T, Hozawa S, Hibi T, Saito I. Prediction of metabolic syndrome using artificial neural network system based on clinical data including insulin resistance index and serum adiponectin. Comput Biol Med. 2011 Nov;41(11):1051-6. doi: 10.1016/j.compbiomed.2011.09.005.
Levin JaI, Eligulashvili TS, Posohov SI, Kovrov GV, Bashmakov MJu. Pharmacotherapy for insomnia: the role of imovan. In: Aleksandrovskii IuA, Vein AM, editors. Rasstroistva sna [Sleep Disorders]. SPb: MIA; 1995. 56-61 pp. (in Russian).
World Medical Association (WMA). WMA Declaration of Helsinki - Ethical Principles for Medical Research Involving Human Subjects. Available from: https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/. Accessed: July 9, 2018.
Viktorova LN, Gorodetskii VK. Colorimetric method for determination of non-enzymatically-glycosylated albumin and hemoglobin. Laboratornoe delo. 1990;(50):15-18. (in Russian).
Zubkov GV, Petrishin VD, Chipirenko VA, Anis'kina AA. Method for determination of melatonin (N-acetyl-5-methoxytryptamine) in urine. Sbornik nauchnykh trudov Kharkovskogo meditsinskogo instituta. 1974;(109):77. (in Russian).
Kulinskii VI, Kostiukovskaia AS. Determination of serotonin in whole blood of humans and laboratory animals. Laboratornoe delo. 1969;(7):390-394. (in Russian).
How to Cite
Copyright (c) 2019 E.V. Misyura, K.G. Manska
This work is licensed under a Creative Commons Attribution 4.0 International License.