ASSESSMENT OF HYPERTENSIVE PATIENTS’ COMPLEX METABOLIC STATUS IN THE NORTHEASTERN REGION OF HUNGARY USING DATA MINING METHODS

Abstract

Objective: Previously, the metabolic status of Central and Eastern European hypertensive adult patients has not been thoroughly explored. Therefore, we aimed to identify the prevalence of hypertension and its associated metabolic abnormalities in a relatively large adult population using data mining methods. Design and method: We analyzed the data of adults who visited the University of Debrecen Clinical Center’s hospital (n = 937,249) by evaluating a period of 20 years (2001-2021). Results: We identified a total of 292,561 adults with hypertension (male/female: 45.4/54.6%), with a 31.2% calculated prevalence of hypertension in the whole study population and a higher prevalence in males (33%) compared to females (29.9%). We found an increasing prevalence in consecutive age groups both in males and females up to the age of 80. The ratio of males with hypertension was higher from 18 to 65 years of age. Markedly higher body mass index (BMI) values were found in hypertensive patients as compared to non-hypertensives (29.5±6.3 vs. 26.57±5.6 kg/m2, p<0.001). A significantly higher triglyceride level was found in hypertensive adults compared to non-hypertensives (1.46 (1.06–2.07) vs. 1.19 (0.82–1.73 mmol/L, p<0.001). We could not find significant differences in total cholesterol and LDL-C levels in the entire patient population (5.1±1.2 vs. 5.0±1.2 mmol/L and 3.1±0.99 vs. 3.1±1.0 mmol/L, respectively). Significantly higher serum glucose levels (5.8 (5.2–6.9) vs. 4.7 (4.6–5.8) mmol/L, p<0.001) were found in subjects with hypertension. Significantly higher urea and creatinine levels were found in hypertensive patients compared to normotensives (5.78 (4.6–7.5) vs. 4.6 (3.7–5.7) mmol/L, p<0.001 and 76 (64–92) vs. 69 (59–81) μmol/L, p<0.001, respectively). Furthermore, we also assessed the rate of significant concomitant diseases in our hypertensive population. Conclusions: The results proved that data mining is an excellent yet not-too-widespread method used to define prevalence in large patient cohorts. This novel way of statistical analysis could be beneficial not only for regional clinical practice but nationwide strategies as well.