• Journal of the Korean Statistical Society
• /
• v.16 no.2
• /
• pp.92-101
• /
• 1987

Our interest is to access in some way teh relative odds or probability that a multivariate observation Z belongs to one of k multivariate normal populations with unequal covariance matrices. We derived the empirical Bayes posterior odds ratio for the classification rule when population parameters are unknown. It is a generalization of the posterior odds ratio suggested by Gelsser (1964). The classification rule does not have complicated distribution theory which a large variety of techniques from the sampling viewpoint have. The proposed posterior odds ratio is compared to the Gelsser's posterior odds ratio through a Monte Carlo study. The results show that the empiricla Bayes posterior odds ratio, in general, performs better than the Gelsser's. Especially, for large dimension of Z and small training sample, the performance is prominent.

• Journal of Preventive Medicine and Public Health
• /
• v.26 no.4 s.44
• /
• pp.628-649
• /
• 1993

This study was conducted to investigate the factors related to the poor school performance of the elementary school children. Two schools in Taegu, one in the affluent area and the other in the poor area, were selected and a total of 175 children whose school performance was within low 10 percentile (poor performers) and 97 children whose school performance were within high 5 percentile (good performers) in each class of 2nd, 4th and 6th grades were tested for the physical health, behavioral problem and family background. Each child had gone through a battery of tests including visual and hearing acuity, anthropometry (body weight, height, head circumference), intelligence (Kodae Stanford-Binet test), test anxiety (TAI-K), neurologic examination by a developmental pediatrician and heavy metal content (Pb, Cd, Zn) in hair by atomic absorption spectrophotometry. A questionnaire was administered to the mothers for prenatal and prenatal courses of the child, family environment, child's developmental history, and child's behavioral and learning problems. Another questionnaire was administered to the teachers of the children for the child's family background, arithmatic & language abilities and behavioral problem. The poor school performance had a significant correlation with male gender, high birth order, broken home, low educational and occupational levels of parents, visual problem, high test anxiety score, attention deficit hyperactivity disorder (ADHD), poor physical growth (weight, height, head circumference) and low I.Q. score. The factors that had a significant correlation with the poor school performance in multiple logistic regression analysis were child's birth order (odds ratio=2.06), male gender(odds ratio=5.91), broken home(odds ratio=9.29), test anxiety score(odds ratio=1.07), ADHD (odds ratio=9.67), I.Q. score (odds ratio=0.85) and height less than Korean standard mean-1S.D.(odds ratio=11.12). The heavy metal contents in hair did not show any significant correlation with poor school performance. However the lead and cadmium contents were high in males than in females. The lead content was negatively correlated with child's grade(P<0.05) and zinc was positively correlated with grade (P<0.05). among the factors that showed a significant correlation with the poor school performance, high birth order, short stature and ADHD may be modified by a good family planning, good feeding practice for infant and child, and early detection and treatment of ADHD. Also, teacher and parents should restrain themselves from inducing excessive test anxiety by forcing the child to study and over-expecting beyond the child's intellectual capability.

• Research in Community and Public Health Nursing
• /
• v.9 no.2
• /
• pp.502-517
• /
• 1998

This research intends to survey family environment, health behavior and health status of the students in urban-rural elementary schools and analyze those factors comparatively, and use the result as basic material for school health teacher to teach health education in connection with family and regional areas. It also intends to improve a pupil's self-abilitiy in health care. The subjects involve 2,774 students of urban elementary schools and 583 student in rural ones, who were selected by means of a multi -stage probability sampling. Using the questionnaire and school documents, we collected data on family environment, health behavior and health status for 19 days. Feb. 2nd 1998 through Feb. 20th 1998. The R -form of Family Environment Scale (Moos, 1974) was used in the analysis of family environment(Cronbach's Alpha =0.80). Questionnaires of Health Behavior in School-aged children used by the WHO in Europe(Aaro et al., 1986) and the ones developed by the Health Promotion Committee of the Western Pacific(WHO, 1995)(adapted by long Young-suk and Moon Young-hee(1996)) were used in the analysis of health behavior, as well documents on absences due to sickness, school health room-visits, levels of physical strength, height, weight and degree of obesity were used to determine health status. In next step, We used them with an $X^2$-test, t-test, Odds Ratio, and a 95% Confidence Interval. 1. In two dimensions of three, family-relationship (t=3.41, p=0.001) and system -maintenances(t= 2.41, p=0.0l6) the mean score of urban children were significantly higher than those of rural ones. In the personal development dimension however, there was little significant difference. Assorting family environment into 10 sub-fields and analyzing them, we recognized that urban children were superior to rural children in the sub-fields of expressiveness (t =3.47, p=0.001), conflict (t=0.48, p=0.001), active-recreational orientation (t = 1.97, p=0.049) and organization (t=4.33, p=0.000). 2. Referring to the Odds Ratios of urban-rural children's health behaviors, urban children set up more desirable behavior than rural children wear ing safety belts (Odds Ratio =0.32, p=0.000), washing hands after meals(Odds Ratio = 0.43, p= 0.000), washing hands after excreting (Odds Ratio = 0.39, p=O.OOO), washing hands after coming - home ( Odds Ratio = 0.75, p = 0.003), brushing teeth before sleeping(Odds Ratio =0.45, p=0.000), brushing teeth more than once a day (Odds Ratio =0.73, p=0.0l2), drinking boiled water (Odds Ratio = 0.49, p=0.000), collecting garbage at home(Odds Ratio=0.31, p=0.000) and in the school(Odds Ratio =0. 67, p=0.000). All these led to significant differences. As to taking milk(Odds Ratio = 1.50, p=0.000), taking care of eyesight(Odds Ratio=1.41, p=0.001) and getting physical exercise in(Odds Ratio = 1.33, p=0.0l9) and outside the school(Odds Ratio = 1.32, p=0.005), rural children had more desirable behavior which also revealed a significant difference. There was little significant difference in smoking, but the smoking rate of rural children(5.5%) was larger than that of urban children(3.9%). 3. Health status was analyzed in terms of absences, school health room-visits, levels of physical strength, and the degree of obesity, height and weight. Considering Odds Ratios of the health status of urban-rural children, the health status of rural children was significantly better than that of the urban ones in the level of physical strength(t=1.51, p=0.000) and the degree of obesity(t=1.84, p=0.000). The mean height of urban children ($150.4{\pm}7.5cm$) is taller than that of their counterparts($149.5{\pm}7.9$), which revealed a significant difference (t =2.47, p=0.0l4). The mean weight of urban children($42.9{\pm}8.6kg$) is larger than that of their counterparts($41.8{\pm}9.0kg$), which was also a significant difference(t=2.81, p=0.005). Considering the results above, we can recognize that there are significant differences in family environment, health behavior, and health status in urban-rural children. These results also suggestion ideas for health education. What we would suggest for the health program of elementary schools is that school health teachers should play an active role in promoting the need and importance of health education, develop the appropriate programs which correspond to the regional characteristics, and incorporate them into schools to improve children's ability to manage their own health management.

• The Journal of Internal Korean Medicine
• /
• v.22 no.3
• /
• pp.423-430
• /
• 2001

Objects: The purpose of this case-control study was done to examine the relationship between stroke prevalence and the anthropometric indices(body mass index(BMI), modified Broca's method(Broca's index), waist circumference(WC), waist-hip ratio(WHR)), dyslipidemia, past medical history(heart disease, hypertension, diabetes mellitus(DM)), life-style factors(smoking, drinking) in Korean adults. Methods : The study group consisted of neurologically confirmed 116 stroke patients as the case group and 116 non-stroke patients as control group. Obesity were defined as $BMI{\geq}25kg/m^2,\;Broca's\;index{\geq}30%$, WC >94cm in male and WC >80cm in female, $WHR{\geq}0.9$ in male and $WHR{\geq}0.8$ in female. Oyslipidemia were defined as total $cholesterol{\geq}200mg/dl,\;triglyceride{\geq}200mg/dl,\;HDL-cholesterol {\leq}35mg/dl,\;LDL-cholesterol{\geq}160mg/dl$. Information on life-style factors and past medical history was obtained from personal interview. The analysis of the data was done by means of chi-square test(Pearson's chi-square test, Fisher's exact test) and student t-test. Results: The results were as follow. In the study group: Hypertension had a 4:05 odds ratio, Broca's $index{\geq}30%$ had a 1.98 odds ratio, WC >94cm in male had a 2.17 odds ratio, WC >80cm in female had a 2.80 odds ratio, $WHR{\geq}0.9$ in male had a 4.66 odds ratio, $WHR{\geq}0.8$ in female had a 5.35 odds ratio, but heart disease, DM, life-style factors, serum lipid and 8MI had no direct relationship with odds ratio(nonsignificant). By student t-test, risk factors for stroke were found to be total cholesterol(p=0,025), LDL-cholesterol(p=0.013), WC(p=0.000) and WHR(p=0.000). Conclusions: This study suggests that people should be advised to control hyperlitension, hyperlipidemia and obesity since these carry a risk of stroke.

• Asian-Australasian Journal of Animal Sciences
• /
• v.18 no.3
• /
• pp.308-313
• /
• 2005

Logistic Regression Analysis was used to compute the odds ratio (OR) and probability of conception of Holstein dairy cows of AL-MARAIE company. Data consisted of 103,778 reproductive records collected from three farms in the central region in the Kingdom of Saudi Arabia. Records were classified according to lactation number, season of calving and milk level. At first lactation, OR of first service was 0.63 of other services and probability of pregnancy from first services was 0.39. Odds ratio increased to 1.72 at fourth lactation or probability of conception reached 0.63. The probability of conception increased from 0.39 for cows inseminated at first services to 0.75 at fifth service insemination. Odds ratio of cows calving in winter were higher than those calving in summer. Odds ratio of conception of low producing cows is about twice as likely to occur from first service as from other services. However, OR of conception of high producing cows was higher than one (probability=0.56) from first service and increased to 1.63 (probability=0.65) from third service. Odds ratio was in favor of the right uterus horn where probability of conception from first service was slightly greater than 50% in first and second lactations and less than 50% in favor of left horn in later lactations.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.9 no.3
• /
• pp.59-65
• /
• 2009

To investigate a posttraumatic stress, social support and work burden and to identify high risk group and related factors which were exerted influence on posttraumatic stress of firefighters. Data were gathered from total 264 firefighters consisting of 85 fire distinguishers & rescue workers, 96 emergency medical personnel and 83 in ambulance & fire truck drivers in C province and were analyzed 22 items by IES-R, 20 work burden items by Choi(2000), 8 social support items by Oh(2006) using SPSSWIN 14.0 program. Posttraumatic stress of general characteristics by work was showed significant difference in age, marriage status, education, position, office duration, mobilization frequency and experience of traumatic events. Posttraumatic stress were showed significant difference in 3 groups and fire distinguishers & rescue workers is the highest group. Posttraumatic stress was correlated with work burden(r=.317, p<0.01) for fire distinguishers & rescue workers, social support(r=-.331, p<0.01) and work burden(r=.522 p<0.001) for emergency medical personnel and work burden(r=.454, p<0.01) for ambulance & fire truck drivers. The high risk groups are fire distinguishers & rescue workers(60%), emergency medical personnel(55.2%)and ambulance & fire truck drivers(45.8%). Related factors are the experience seeing victim's danger for fire distinguishers & rescue workers(odds ratio=1.216, 95% confidence interval:1.068-1.383), and are work burden(odds ratio=1.100, 95% confidence interval:1.043-1.159) and office duration(odds ratio=1.010, 95% confidence interval 1.001-1.018) for emergency medical personnel, and are the experience seeing victim's danger(odds ratio=1.178, 95% confidence interval:1.010-1.373), age(odds ratio=1.129, 95%confidence interval:1.020-1.249), work burden(odds ratio=1.103, 95% confidence interval:1.034-1.177) for ambulance & fire truck drivers.

• The Journal of Korean Society for School & Community Health Education
• /
• v.16 no.2
• /
• pp.17-31
• /
• 2015

Objectives: This study aimed to analyze the relationship between metabolic syndrome and cataract in a middle-aged men over the age of 40's by using the data from the 5th Korean National Health and Nutrition Examination Survey(2010~2012). Methods: The subjects for the analysis were comprised of 5,808 patients who were at least 40 years or more with metabolic syndrome and cataract. The comparison between the metabolic syndrome and ophthalmic diseases according to the health demographic characteristics, health behavior and health conditions was analyzed with ${\chi}^2-test$. The complex samples logistic regression analysis was used to calculate the odds ratio of cataract according to the health demographic characteristics, health behavior and health conditions and the combination of the comprising factors of the metabolic syndrome. Results: After controlling the confounding factors, the odds ratio of cataract in the metabolic syndrome group was OR 1.30, which was significantly high (p<0.01). Furthermore, the odds ratio of cataract was significantly high in the case of males, as the age increased, educational level decreased, with exercising habits of less than 3 days a week, sun exposure of at least 5 hours, with hypertension and diabetes (p<0.01). The results of analyzing the odds ratio for the occurrence of cataract with the components of metabolic syndrome were OR 1.36 (concurrent high fasting blood glucose, high triglycerides, low HDL cholesterol), OR 1.63 (concurrent high fasting blood glucose, high triglycerides, high waist circumference) and OR 1.33 (high triglycerides, low HDL cholesterol, high waist circumference), which showed significantly high odds ratio for cataract occurrence (p<0.05). Conclusions: We come to know that the relation between metabolic syndrome and components are associate with cataract. Therefore the thorough management of metabolic syndrome and components is needed to prevent cataract.

• The Korean Journal of Applied Statistics
• /
• v.18 no.2
• /
• pp.471-485
• /
• 2005

For categorical data analysis, the collapsibility were explained with the odds ratio (cross-product ratio). When these theories with these odds ratios are applied to real $2{\times}2{\times}K$ contingency tables, it is impossible to decide whether data are collapsible. Among graphical methods to represent odds ratios, Contour plot which is developed by Doi, Nakamura and Yamamoto (2001) could explain the structure of these data, but cannot decide on the collapsibility. In this paper, by using the Raindrop plot proposed by Barrowman and Myers (2003), we suggest an alternative method which can not only explain the structure of data, but also decide on the collapsibility.

• Journal of the Korean Dietetic Association
• /
• v.25 no.2
• /
• pp.83-104
• /
• 2019

Elevated high-sensitivity C reactive protein (hs-CRP) levels and metabolic syndrome are considered important predictors of cardiovascular mortality. This study examined the influence of the alcohol consumption level on the hs-CRP level and the prevalence of metabolic syndrome in South Koreans. The study subjects were 3,884 participants (${\geq}19$ years) from the KNHANES in 2015. The subjects were divided into four groups according to their alcohol consumption; None (none-alcohol consumption), Low (less than 15 g/day), Medium (15~29.9 g/day), and High (over than 30 g/day). The odds ratio for high-risk hs-CRP of men was significantly lower in the Low or Medium groups compared to the None group. In women, the odds ratio for high-risk hs-CRP was not decreased in the Low or Medium groups but was significantly higher in the High group than in the None group. Compared to nondrinkers, the prevalence of metabolic syndrome was lower in the Low group and Medium group in both men and women, whereas it was significantly higher in the High group in men only. In particular, the prevalence of low HDL-cholesterol was lower in all drinking groups. When the confounding factors were adjusted, the odds ratio for low HDL-cholesterol appeared to decline in the Low groups in both men and women. On the other hand, in the Medium group, women showed an elevated odds ratios of high blood pressure (HBP) and high fasting plasma glucose but men showed an elevated odds ratios of HBP and high triglyceride levels. The overall results suggest that low alcohol consumption is more appropriate than medium alcohol consumption. Nevertheless, more study will be needed to evaluate the appropriate alcohol consumption level.

• Journal of The Korean Society of Clinical Toxicology
• /
• v.15 no.2
• /
• pp.122-130
• /
• 2017

Purpose: To evaluate the effects of low-dose intravenous N-acetylcysteine on the prevention of contrast-induced nephropathy (CIN) in patients undergoing computed tomography (CT). Methods: All patients presenting to our emergency department and undergoing CT with intravenous contrast media between August 2014 and April 2016 were retrospectively enrolled. We included hospitalized patients with renal dysfunction [estimated glomerular filtration rate (GFR) between 30 and $89mL/min/1.73m^2$]. A 600-mg injection of N-acetylcysteine was given to patients once before and once immediately after CT, depending on the preference of physician. The primary outcome was CIN defined as an increase in creatinine level of ${\geq}25%$ or ${\geq}0.5mg/dL$ from the baseline within 48 to 72 hours after CT. A trained person blindly reviewed all medical records. Results: Of the 1903 admitted patients, CIN occurred in 9.8% of patients who received 1200 mg intravenous N-acetylcysteine (24/244) and 6.8% of patients who did not (113/1659, p=0.090). In a multivariable regression analysis, N-acetylcystine was not relevant to the prevention of CIN (odds ratio=1.42 [95% CI, 0.90-2.26]). Even in the stratified analysis using the propensity score matching, N-acetylcysteine was irrelevant (GFR 30-59: odds ratio=1.06 [95% CI, 0.43-2.60]; GFR 60-89: odds ratio=1.76 [95% CI, 0.75-4.14]). After adjustment, crystalloids were significantly associated with the reduction in CIN compared with dextrose water (odds ratio=0.60 [95% CI, 0.37-0.97]). Conclusion: No effect was found when low-dose intravenous N-acetylcysteine was used to prevent CIN. However, there seems to be an association between crystalloids and reduction in CIN.