Neural networks and logistic regression: Analysis of a case-control study on myocardial infarction

Objectives: To explore the association between single-nucleotide polymorphisms (SNPs) in MTHFR and APOE and the risk of CAD and, more importantly, the severity of CAD and the profile of serum lipids, we performed a case-control study in a Chinese Han population. Methods: A total of 1,207 cases of consecutive CAD-suspected inpatients were recruited, and 406 CAD cases and 231 non-CAD controls were enrolled for the final analysis after screening for exclusion criteria. All subjects had undergone coronary angiography, and the severity of CAD was evaluated by 2 cardiologists according to the Gensini scores. The genotypes of MTHFR and APOEwere detected using real-time PCR, and then verified by Sanger sequencing. Environmental risk factors, such as age, sex, smoking, alcohol consumption, hypertension, diabetes, dyslipidemia, and BMI were collected. Statistical analyses (the χ2 test, binary logistic regression analysis, and ordinal polytomous logistic regression analysis) were performed with SPSS v16.0. Results: The genotypes ofall the subjects included in the CAD and non-CAD groups in this study were successfully detected, with an agreement of 100% with Sanger sequencing. The distributions of genotypes CT and TT at MTHFR C667T were higher in CAD cases than in non-CAD controls (OR 1.99, 95% CI 1.34–2.95; OR 1.77, 95% CI 1.18–2.67; p < 0.05), whereas genotype AC at MTHFR A1298Cwas lower in CAD cases (OR 0.71, 95% CI 0.50–1.02; p < 0.05). A significant association was observed in genotypes CT and TT at MTHFR C667T and the risk of CAD (OR 1.44, 95% CI 1.27–3.67; OR 1.56, 95% CI 0.88–2.78; p < 0.05). Both genotypes and alleles of APOE were comparable in the CAD cases and non-CAD controls (p > 0.05). The genotype TT at MTHFR C667T and ε4+ at APOE were more likely to be found in the CAD subgroup with a Gensini score ≥72 (p = 0.040 and p = 0.028, respectively). Meanwhile, in the patients with genotype TT,a higher level of serum Hcy was detected, while genotype ε4+ patients possessed higher levels of serum apolipoprotein E (ApoE) and low-density lipoprotein cholesterol (LDL-C) than other genotypes. Conclusion: This study revealed that the SNP site of MTHFR C667Tis associatedwith the risk of CAD in this Chinese Han population. In addition, the genotypes of TT in MTHFR C667T and ε4+in APOE may increase the severity of CAD, and higher Hcy, LDL-C, and ApoE levels may be involved in this pathogenic process.

Download Full-text

Malignant melanoma risk factors by anatomic site: A case-control study and polychotomous logistic regression analysis

International Journal of Cancer ◽

10.1002/(sici)1097-0215(19960904)67:5<636::aid-ijc8>3.0.co;2-v ◽

1996 ◽

Vol 67 (5) ◽

pp. 636-643 ◽

Cited By ~ 36

Author(s):

Ya-Ting Chen ◽

Robert Dubrow ◽

Theodore R. Holford ◽

Tongzhang Zheng ◽

Raymond L. Barnhill ◽

...

Keyword(s):

Risk Factors ◽

Logistic Regression ◽

Regression Analysis ◽

Malignant Melanoma ◽

Logistic Regression Analysis ◽

Case Control Study ◽

Case Control ◽

Melanoma Risk ◽

Anatomic Site ◽

Control Study

Download Full-text

Determinants of impaired growth among hospitalized children: a case-control study

Sao Paulo Medical Journal ◽

10.1590/s1516-31802004000300008 ◽

2004 ◽

Vol 122 (3) ◽

pp. 117-123 ◽

Cited By ~ 10

Author(s):

Marilia de Carvalho Lima ◽

Maria Eugênia Farias Almeida Motta ◽

Eliane Cavalcanti Santos ◽

Gisélia Alves Pontes da Silva

Keyword(s):

Risk Factors ◽

Logistic Regression ◽

Regression Analysis ◽

Logistic Regression Analysis ◽

Low Income ◽

Case Control Study ◽

Protein Energy Malnutrition ◽

Case Control ◽

Protein Energy ◽

Control Study

CONTEXT: Protein energy malnutrition constitutes a public health problem, especially in less affluent countries. The identification of amenable predictive risk factors is of major importance for policy makers to plan interventions to reduce infant malnutrition. OBJECTIVE: To identify risk factors for protein energy malnutrition among hospitalized low-income children aged 6 to 24 months. TYPE OF STUDY: Case-control study. SETTING: Two public hospitals in Recife, Brazil. PARTICIPANTS: The cases were 124 infants with length-for-age below the 10th percentile of the National Center for Health Statistics curve and the controls were 241 infants with length-for-age equal to or above the 10th percentile who were recruited in the same infirmary. METHODS: Cases and controls were compared in relation to a variety of sociodemographic, environmental and reproductive factors, and their healthcare, previous feeding practice and morbidity. Logistic regression analysis was used to investigate the net effect of risk factors on infant malnutrition, after adjusting for potential confounding variables. RESULTS: The mother's age, possession of a TV set, type of water supply, family size and location of the home were significantly associated with child malnutrition in the bivariate analysis. However, these associations lost their significance after adjusting for other explanatory variables in the hierarchical logistic regression analysis. This analysis showed that low birth weight contributed the largest risk for impaired growth. Increased risks of infant malnutrition were also significantly associated with households that had no toilet facilities or refrigerator, high parity for the mother, no breastfeeding of the infant, inadequate vaccination coverage and previous hospitalization for diarrhea and pneumonia. DISCUSSION: The literature shows that chronic malnutrition, as assessed by low length-for-age indexes, is often related to low income. However, this was not the case in this study, in which other variables had greater impact on child growth. CONCLUSIONS: In view of the multiple causes of malnutrition, the interrelationship among its determinants should be taken into account when adopting strategies for its reduction and prevention.

Download Full-text

Risk Factors for Low Birth Weight in Sidama Zone Government Hospitals, Southern Ethiopia, A Case-control Study

10.21203/rs.3.rs-288190/v1 ◽

2021 ◽

Author(s):

Temesgen Tilahun ◽

Hailu Hailemariam

Keyword(s):

Risk Factors ◽

Logistic Regression ◽

Regression Analysis ◽

Birth Weight ◽

Low Birth Weight ◽

Logistic Regression Analysis ◽

Case Control Study ◽

Case Control ◽

Multivariable Logistic Regression Analysis ◽

Control Study

Abstract Background: Many studies had been conducted on the epidemiology of low birth weight (LBW) in high-income countries, however in countries like Ethiopia such evidence is scarce. The objective of this study was to assess the risk factors for LBW in Sidama zone.Method: Hospital-based case-control study design with a total sample size of 480, 96 cases, and 384controls from March to June 2018. Newborns were selected from three hospitals in the zone using simple random sampling techniques. Data were collected using a pretested questionnaire. Anthropometric measurements were made by following standard procedure. Risk factors for LBW were identified using multivariable logistic regression analysis. The output is presented using an adjusted odds ratio (AOR) with a 95% confidence interval (CI). Result: Multivariable logistic regression analysis shows that urban place of residence [AOR=2.55(95% CI=1.15-5.82], mothers who did not have iron supplement during pregnancy [AOR=12.5(95 CI=5.06-31.0)], premature birth [AOR=2.78 (95% CI= (1.27-6.06)] and history of pregnancy complication[AOR=7.60(2.03-28.45)] were found to be significant determinants of LBW. Conclusion: Since the prevalence of LBW was more prominent in rural areas than their urban counterparts, socio-economic empowerment of rural women and community-based nutrition promotions programs should be given special emphasis. Strengthening efforts on availing basic health services and promoting education on nutrition during pregnancy also have positive inputs. Iron supplementation and nutritional assessment of women during ANC follow-up and providing interventions accordingly should be strengthening in all health facilities.

Download Full-text

Alzheimer’s Disease and its Risk Factors: A Case-control Study

Trends in Medical Sciences ◽

10.5812/tms.113857 ◽

2021 ◽

Vol 1 (1) ◽

Author(s):

Mojtaba Dolatshahi ◽

Sara Mahmoudi ◽

Mansour Torabi nia ◽

Vahid Kheirandish

Keyword(s):

Risk Factors ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Logistic Regression ◽

Regression Analysis ◽

Chronic Disease ◽

Logistic Regression Analysis ◽

Case Control Study ◽

Case Control ◽

Control Study

Objectives: The etiology of Alzheimer’s disease is supposed to involve environmental exposure and genetic vulnerability. The present study aimed to assess the association between Alzheimer’s disease and its risk factors. Methods: We conducted a case-control study of 95 Alzheimer’s disease patients and 98 matched controls. All participants (case and control groups) were examined by mini-mental state examination. This information was collected by a risk factor questionnaire from January to June 2019. Data were analyzed using logistic regression analysis via SPSS-22 software. Results: The findings revealed the effect of risk factors' odds ratios on the occurrence of Alzheimer’s disease by logistic regression analysis. Sex (female), chronic disease, loneliness or being single, and family history showed positive associations with AD, whereas daily physical exercise, religious beliefs, high level of social communication, and male sex had negative associations with AD development (P < 0.05). Conclusions: The study highlighted the difficulty of etiology and recommended that the effective interventions for social support of older people, psychological condition, chronic disease, and lifestyle may be promising preventive policies.

Download Full-text

Siesta and Risk for Ischemic Stroke: Results from a Case-Control Study

Medicina ◽

10.3390/medicina56050222 ◽

2020 ◽

Vol 56 (5) ◽

pp. 222

Author(s):

Yousef Mohammad

Keyword(s):

Logistic Regression ◽

Regression Analysis ◽

Ischemic Stroke ◽

Protective Effect ◽

Case Control Study ◽

Multivariate Logistic Regression Analysis ◽

Case Control ◽

Multivariate Logistic Regression ◽

Neurology Clinic ◽

Control Study

Background and objectives: Siesta, which is a short afternoon nap, is a habit that is commonly practiced in the Mediterranean and tropical areas. Data on the association between siesta and coronary artery disease has been conflicting. A protective effect has been demonstrated in the countries that commonly practice siesta, but a harmful effect has been observed in the countries that infrequently practice the siesta habit. Information on the association between siesta and ischemic stroke has been, however, lacking. Hence, the purpose of our study was to determine the effect of siesta on ischemic stroke. Materials and Methods: This was a case-control study, conducted on the patients with acute ischemic stroke who came for their first follow-up visit to the neurology clinic. Controls were randomly selected from the patients visiting the neurology clinic on the same day as the patients with ischemic stroke. In addition to basic demographics and the occurrence of established stroke risk factors, information about siesta practice was also collected from both groups. A multivariate logistic regression analysis was utilized to determine the relationship between siesta practice and ischemic stroke. Results: A total of 206 patients were recruited from the neurology clinic of King Khalid university hospital; of which only 194 subjects were included in the analysis (98 ischemic stroke cases and 96 controls). The mean age of the participants was 59.68 ± 13.75 years and 98 (50.52%) were male. Interestingly, 43% of the whole study cohort practiced regular siesta. However, when compared to the stroke population, the control group practiced siesta more frequently (30% vs. 56%). In a multivariate logistic regression analysis, hypertension, diabetes mellitus, excess body weight (body mass index > 25 kg/m2) and dyslipidemia were found to increase the risk of ischemic stroke (OR 2.12, 95% CI: 1.02–4.66, p = 0.005; OR 2.72, 95% CI: 1.94–4.88, p = 0.014; OR 2.94, 95% CI: 1.5164–5.7121 p = 0.0014; OR 3.27, 95% CI: 2.42–5.199, p ≤ 0.001, respectively). On the contrary, the practice of regular siesta lowered the risk of ischemic stroke (OR 0.58, 95% CI: 0.3551–0.9526, p = 0.031). Conclusions: Siesta was associated with a reduced risk for the occurrence of ischemic stroke. Large prospective longitudinal studies should be conducted to verify the protective effect of siesta on stroke.

Download Full-text

Determinants of postnatal care non-utilization among women in Demba Gofa rural district, southern Ethiopia: a community-based unmatched case-control study

BMC Pregnancy and Childbirth ◽

10.1186/s12884-020-03244-9 ◽

2020 ◽

Vol 20 (1) ◽

Author(s):

Markos Manote ◽

Tsegaye Gebremedhin

Keyword(s):

Logistic Regression ◽

Regression Analysis ◽

Postnatal Care ◽

Case Control Study ◽

Care Service ◽

Case Control ◽

Rural District ◽

Southern Ethiopia ◽

Community Based ◽

Control Study

Abstract Background Maternal and neonatal mortality remain a significant problem across much of the developing world, especially in sub-Saharan Africa countries. In Ethiopia, most maternal and neonatal deaths occur during the postpartum period; this is a critical time for monitoring the health of women and newborns, but the most neglected period for care. In rural communities of Ethiopia, the utilization of postnatal care service is very low and evidence on which factors contribute to the non-utilization of postnatal care (PNC) is insufficient. Consequently, this study was designed to identify the determinants of postnatal service non-utilization among women who gave birth in Demba Gofa rural district, Southern Ethiopia. Methods A community-based unmatched case-control study was conducted among 186 cases (postnatal care non-utilizers) and 186 controls (postnatal care utilizers) in Demba Gofa rural district from March 1 to April 10, 2019. A previously tested interviewer-administered structured questionnaire was used for data collection. Binary logistic regression analysis was performed. In the final multivariable logistic regression analysis model, a p-value of less than 0.05 and an Adjusted Odd Ratio (AOR) with a 95% confidence interval (CI) was used to determine variables for postnatal care non-utilization. Results In this study, women who delivered recently were incorporated within 186 cases and 186 controls. Not knowing the availability of PNC services (AOR: 4.33, 95% CI: 1.71–10.99), having a home delivery (AOR: 7.06, 95% CI: 3.71–13.44), ANC non-attendance (AOR: 6.14, 95% CI: 3.01–12.50), unable to make an independent decision (AOR: 9.31, 95% CI: 3.29–26.35), and not participating in the Women’s Development Army (WDA) (AOR: 5.09, 95% CI: 2.73–9.53) comprised the determinants which were assessed for non-utilization of postnatal care services. Conclusions Encouraging institutional delivery along with integrated health education about postnatal care and postnatal danger signs, empowering women to execute independent decisions, accessing PNC services and strengthening participation in the Model Families will likely improve postnatal care service utilization in the district of Ethiopia.

Download Full-text

Maternal Age and Parity as Predictors of Human Twinning

Acta geneticae medicae et gemellologiae twin research ◽

10.1017/s0001566000005237 ◽

1990 ◽

Vol 39 (3) ◽

pp. 329-334 ◽

Cited By ~ 13

Author(s):

B. Bønnelykke

Keyword(s):

Logistic Regression ◽

Regression Analysis ◽

Random Sample ◽

Maternal Age ◽

Case Control Study ◽

Positive Association ◽

Case Control ◽

Negative Association ◽

Medical Birth Register ◽

Control Study

AbstractA case-control study was conducted to evaluate the association between maternal age and parity and dizygotic (DZ) and monozygotic (MZ) twinning, respectively. Mothers of all twins born alive in Denmark in 1984 or 1985 were included into the study as cases, and a random sample of mothers of singletons were controls. Data on maternal age at delivery and number of previous births were collected from the Danish Medical Birth Register. By means of logistic regression analysis, a significant and positive association was found betwen maternal age and DZ twinning, and a significant and negative association between parity and MZ twinning. No association was found for parity and DZ twinning, nor for maternal age and MZ twinning. The study suggests that human MZ twinning has predictors too, as has DZ twinning. The finding has implications for zygosity classification in future twin research.

Download Full-text

Specific Human Papillomavirus Types and Other Factors on the Risk of Cervical Intraepithelial Neoplasia: A Case-Control Study in Korea

International Journal of Gynecological Cancer ◽

10.1111/igc.0b013e3181e4a720 ◽

2010 ◽

Vol 20 (6) ◽

pp. 1067-1073 ◽

Cited By ~ 7

Author(s):

Chan Joo Kim ◽

Young Suk Lee ◽

Hyun Sung Kwack ◽

Won Sik Yoon ◽

Tae Chul Park ◽

...

Keyword(s):

Logistic Regression ◽

Regression Analysis ◽

Cervical Intraepithelial Neoplasia ◽

Logistic Regression Analysis ◽

Case Control Study ◽

Sexual Debut ◽

Multiple Logistic Regression Analysis ◽

Intraepithelial Neoplasia ◽

Multiple Logistic Regression ◽

Control Study

Objective:Cervical neoplasia is attributed to a persistent human papillomavirus (HPV) infection. We performed a hospital-based, case-control study to evaluate the associations of HPV genotypes and other cofactors with cervical intraepithelial neoplasia (CIN).Methods:A total of 158 women were enrolled after we received their informed consent, and the control group (the non-CIN group; n = 80) was selected from women at St Paul's Health Promotion Center. The CIN group (n = 78) was enrolled from the outpatient clinics at Kangnam St Mary's Hospital. Cervical intraepithelial neoplasia was confirmed with colposcopic-guided biopsy or loop electrosurgical excision procedure-conization. A structured questionnaire, Papanicolaou test, and HPV testing were completed. We compared each risk factor using χ2tests and simple logistic regression analysis between the CIN and non-CIN groups. Finally, odds ratios (ORs) were calculated again by multiple logistic regression analysis.Results:The most frequent HPV types in CIN were HPV-16, HPV-58, HPV-31/-33, and HPV-35/−56. The OR of the A9 group (HPV-16, HPV-31, HPV-33, HPV-35, HPV-52, HPV-58) was 22.7 (95% confidence interval [CI], 8.3-62.5), that of the A6 group (HPV-53, HPV-56, HPV-66) was 2.9 (95% CI, 1.1-7.5), and that of the A7 group (HPV-18, HPV-39, HPV-45, HPV-59, HPV-68) was 1.5. Sexual debut before 20 years old had significantly higher OR than did a sexual debut after 30 years (OR, 32.9; 95% CI, 2.8-364.7). The OR for CIN in single women versus married women was 6.2 (95% CI, 2.5-15.2). Compared with parous women (parity >3), nonparous women had a higher OR (95% CI, 1.4-16.7). On the multiple logistic regression analysis including the sexual debut age, the marital status, parity, cytology, and the HPV groups, the A9 group had a significant OR for CIN (6.1; 95% CI, 1.6-23.6).Conclusions:The risk of CIN was higher for women infected with the HPV-A9 group after multiple logistic regression analysis. The other clinical risk factors were not significant factors of CIN.

Download Full-text

Coffee consumption and the risk of polycystic ovary syndrome: evidence from a case-control study

10.21203/rs.3.rs-889597/v1 ◽

2021 ◽

Author(s):

Yue Wang ◽

Yin Yang ◽

Hui Deng ◽

Xiao-Qing Bu ◽

Tian Li ◽

...

Keyword(s):

Logistic Regression ◽

Regression Analysis ◽

Polycystic Ovary Syndrome ◽

Case Control Study ◽

Polycystic Ovary ◽

Coffee Consumption ◽

Case Control ◽

Confounding Factors ◽

Ovary Syndrome ◽

Control Study

Abstract Background Coffee has become one of the most common beverages worldwide. Some components in coffee have been reported to be beneficial to the regulation of hormones related to polycystic ovary syndrome(PCOS), and therefore may limit the development of PCOS. However, epidemiological evidence on the relationship between coffee consumption and PCOS is limited. In this study, we aimed at investigating the association between coffee consumption and PCOS risk. Methods The study was a hospital-based case-control study conducted at Chongqing Hospital of Traditional Chinese Medicine, China, from July 2018 to January 2020. Data was collected from 206 women with PCOS and 206 healthy women. A self-designed structured questionnaire was employed to collect relevant information from study participants. Sleep quality was evaluated by Pittsburgh Sleep Quality Index (PSQI). Logistic regression analysis was used to calculate odds ratios (ORs) and 95% confidence intervals (95% CI), adjusted for other confounding factors. Results A statistically significant inverse correlation was observed between coffee consumption and PCOS. After adjusting all potential confounding factors, the results of logistic regression analysis showed that compared to non-drinker, the ORs(95%CI) for ≤ 1 cup/week, 2–3 cups/week and > 3 cups/week were 0.322(0.180–0.574), 0.263(0.104–0.664) and 0.152(0.046–0.563), respectively, and the linear trend was also statistically significant(P for trend < 0.01). Conclusions Coffee consumption was found to be associated with a reduced risk of PCOS. Given the scarcity of studies on coffee consumption and PCOS, as well as the inherent limitations of our study, the prospective cohort studies in the future are needed to verify our results.

Download Full-text