Clinical characteristics
Tables 1 and 2 explain the results of the logistic regression test among total clinical characteristics in each gender. There were differences in gender-specific clinical characteristics. Among 1,781 total patients, 977 patients and 804 patients were male and female, respectively. Patients had several disease-association factors in both genders: area, positive diagnosis of hyperlipidemia, positive diagnosis of mental disease, continual use of one or more drugs, waist, and height.
Through previous studies, how these factors are related with gastritis was verified.
The incidence of gastritis is affected by population, geographic variation, or lifestyle [1, 5]. In this study, we used cohort data that comprised country and city populations. As shown Tables 1 and 2, in both genders, the population of local A had a higher gastritis incidence rate. This means that differences in patient lifestyle and environment affect the disease incidence rate.
It is well known that stress has an influence on carcinogenesis. A study reported that gastritis is closely connected with mental illness [12] and differs in degree according to drug use-taking medicine affects the stomach. Nonsteroidal anti-inflammatory drugs can cause gastrointestinal damage [13].
Losing appetite is the one of phenotypes of gastritis that are connected with waist size or hip size as gastritisassociated factors. However, definite evidence is lacking [14]. As shown in Tables 1 and 2, we can explain the associative relation that waist size or hip size is smaller despite patients being taller than normal. Also, height is considered to be a gastritis risk factor. As shown in a previous study, higher height tends to increase the gastric cancer incidence rate [15].
In males and females, hyperlipidemia and gastritis are related with gastritis. Only male cases are associated with vascular disease. Coronary artery has the potential to grow into myocardial infarction [16], and myocardial infarction is related with blood pressure [17]. As previously reported, approximately 30% of patients with coronary heart disease suffer from hyperlipidemia at the same time [18]. So, hyperlipidemia, coronary artery, and myocardial infarction are all related to each other. As shown in Table 1, in males, by logistic regression, there is a relation with hyperlipidemia, coronary artery, myocardial infarction, hyperlipidemia therapy, hypotensive or antihypotensive drug use, and diastolic blood pressure. This relation is more remarkable in males than in females. Repeatedly, it can be considered that vascular disease is connected with gastritis. In fact, in this study, including males and females, 16 patients had gastritis and myocardial infarction (20%), 77 patients had gastritis and hyperlipidemia (36%), and 13 patients had gastritis and coronary artery (25.5%).
Female patients with gastritis have 4 more gastritis-associated factors by logistic regression test compared with males. In the case of taking osteoporosis medicine, the incidence rate is 1.7 times higher compared with normal. Taking anticonvulsants is a meaningful factor by logistic regression test, but few patients were taking this medicine [4]; so, it is hard to prove that taking anticonvulsants has a positive relation with gastritis. Patients who were diagnosed with tuberculosis and patients who have taken tuberculosis drugs suffer gastritis more than normal subjects. This gives information on the relation between gastritis and tuberculosis in females.
ROC curves can show how gastritis-associated factors of Tables 1 and 2 affect gastritis patients and normal subjects. Whether these factors are conclusive should be confirmed by AUC values. Figs. 1 and 2 are ROC curves using the result of the logistic regression test of Tables 1 and 2, respectively. Each AUC value is 0.697 and 0.687. This result means that these factors are not useful as gastric-specific markers.